RoboWiki - User contributions [en]

File:Explore.zip

2025-07-03T11:05:23Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-03T01:45:13Z

Robot:

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
(Note: All included graphs represent point clouds obtained from LiDAR scans relative to the robot's position, which is set at the origin (0,0). The straight lines in the plots correspond to wall approximations)

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[Media:explore.zip|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:26:06Z

Robot:

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
(Note: All included graphs represent point clouds obtained from LiDAR scans relative to the robot's position, which is set at the origin (0,0). The straight lines in the plots correspond to wall approximations)

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:25:47Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
Note: All included graphs represent point clouds obtained from LiDAR scans relative to the robot's position, which is set at the origin (0,0). The straight lines in the plots correspond to wall approximations computed using the RANSAC algorithm.

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:24:23Z

Robot:

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
(comment: all included graphs represent obtained from LiDAR points relative to the robot which position is (0,0) and lines which approximate walls)

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:23:24Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
(comment: all included graphs represent obtained from LiDAR points relative to the robot which position is (0,0)

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:22:58Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:
(comment: all included graphics represent obtained from LiDAR points relative to the robot which position is (0,0)

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:22:42Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found (comment: all included graphics represent obtained from LiDAR points relative to the robot which position is (0,0):

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:15:25Z

Robot: /* Description of the program */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These are considered gaps, typically representing doors. Such information can help refine the robot’s position beyond raw odometry data.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:14:08Z

Robot: /* Description of the program */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if the slope k in the line equation y = kx + b is greater than 0, the robot turns slightly to the right; if it is less than 0, it turns to the left. This approach proved effective in the corridor environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:13:19Z

Robot: /* Description of the project */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the program ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than threshold "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Prototyp Robotickej stavebnice pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:12:34Z

Robot: Created page with "== Cieľ projektu == Cieľom tohto projektu bolo navrhnúť základy prototypu '''robotickej stavebnice''' pre novú kategóriu v '''Robotickej Lige'''. Táto stavebnica sl..."

== Cieľ projektu ==

Cieľom tohto projektu bolo navrhnúť základy prototypu '''robotickej stavebnice''' pre novú kategóriu v '''Robotickej Lige'''.

Táto stavebnica slúži ako '''didaktická pomôcka''', ktorá podporuje rozvoj detskej predstavivosti a súťaživosti v oblasti robotiky.

Inšpiráciou boli existujúce stavebnice ako '''LEGO Mindstorms''', '''Totem''' a '''Merkur'''.

Základom návrhu sú univerzálne stavebné "kocky", ktorých rozmery určujú kompatibilitu všetkých modulov.

== Realizácia projektu ==

=== Vlastnosti stavebnice ===

* '''Modularita''' – stavebnica je prispôsobiteľná rôznym typom úloh
* '''Jednoduchosť''' – vhodná na jednoduché použitie aj pre začiatočníkov
* '''3D tlačiteľnosť''' – všetky komponenty sú navrhnuté na domácu 3D tlač

=== Moduly ===

Vybrané prototypové moduly:

* Základné kocky
* Motor
* Joystick
* Senzor vzdialenosti

== Prototypy ==

=== Základné kocky ===

Základné stavebné prvky majú rozmery založené na štandarde skrutiek M3, s osovou vzdialenosťou 8 mm.
Dizajnovo pripomínajú diely zo stavebnice Merkur.

Kocky sú plne '''3D tlačiteľné''' a na konci dokumentu je uvedený odkaz na všetky modely.

[[File:Zaklad.png|250px]]
[[File:L.png|250px]]

=== Motor ===

Krabička pre servo motor bola navrhnutá tak, aby bola:

* kompaktná
* jednoducho tlačiteľná
* ergonomická

Servo sa vkladá zo spodnej strany krabičky a upevňuje pomocou skrutiek.
Ozubené koleso je navrhnuté tak, aby sa nepretáčalo a mohlo byť v prípade potreby aj priskrutkované.

'''Hlavné vlastnosti:'''
* Možnosť vytlačiť krabičku ako jeden celok
* Pevné uchytenie osi
* Jednoduchá výmena motora v prípade poruchy

[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]

=== Joystick ===

Prototyp umožňuje pevné pripojenie ku "kockám".
Na bokoch sú otvory pre matice M3, ktoré sa nepretáčajú vďaka presnému výrezu.
Vnútorné výstupky zabezpečujú stabilné uloženie joysticku.

'''Hlavné vlastnosti:'''
* Jednoduchosť
* Rýchle osadenie joysticku do krabičky
* Jednoduché pripojenie ku základným kockám

[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]

=== Senzor vzdialenosti ===

Základný a nevyhnutný modul – ultrazvukový senzor vzdialenosti – je navrhnutý tak, aby:

* bol jednoducho integrovaný do stavebnice
* bol odolný voči nárazom
* mal presné úchyty pre senzor a výrezy pre prístup k pinom

'''Hlavné vlastnosti:'''
* Odolná konštrukcia
* Presné uchytenie senzora
* Otvory pre jednoduché pripojenie k stavebnici
* Prístup ku konektorom cez spodnú časť krabičky

[[File:Ultrazvuk final.jpg|250px]]
[[File:Ultrazvuk.png|250px]]

== Nedostatky a možné úpravy ==

=== Motor ===

* Chýba možnosť pripevnenia ku stavebnici
* Nedostatočne stabilná podstava

→ V ďalšej verzii plánujem rozšíriť podstavu a pridať upevňovacie body.

=== Joystick a senzor ===

* Slabé uchytenie vrchného krytu
* Joystick sa pri používaní hýbe

→ Riešenia:
* Nahradiť zasúvací kryt za zacvakávací s drážkami
* Joystick bude priskrutkovaný do krabičky

== Ďalšie možnosti do budúcnosti ==

* Vývoj riadiacej jednotky s čipom '''RP2040-Zero''', napájanou lítiovou batériou
* Jednotka bude riadiť komunikáciu všetkých modulov
* Možnosť rozšírenia o '''gyroskop''' pre sledovanie natočenia robota (napr. pre autíčko)

== Záver ==

Tento projekt položil pevné základy pre vývoj '''robotickej stavebnice pre Robotickú ligu'''.
Modulárnosť, 3D tlač a jednoduchosť robia zo stavebnice skvelý nástroj na učenie robotiky u detí.

Do budúcna plánujem:
* Dokončiť riadiacu jednotku
* Opraviť prototypy podľa identifikovaných nedostatkov
* Rozšíriť počet funkčných modulov

[[File:Komplet.jpg|450px]]

== Video ukážka ==

[[File:Servo video.mp4]]

__NOTOC__

Projects AI Robotics

2025-07-02T21:12:24Z

Robot:

This page contains student projects from the course Algorithms for AI Robotics

Please create a separate page for each project:

== 2025 ==

[[Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich]]

[[Prototyp Robotickej stavebnice pre Robotickú ligu - Sebastián Horváth]]

[[Comparing Reinforcement Learning with Neuroevolution - Tomáš Bisták]]

== 2024 ==
[[Smart Home Model for Optimal Room Temperature and Air Quality - Celina Mueller]]

[[VisionAir: Camera Controlled CoDrone - Anna Karikó & Julia Liem]]

[[2 Projects: Tello Drone Arrow Navigator / Tello Drone Detection of Line Numbers - Ibrahim Hayik]]

[[Spike Prime - Space Invaders - Unax Murua, Eric Ayestaran, Leire Sáez de Cortázar | Space Invaders - Unax Murua, Eric Ayestaran, Leire Sáez de Cortázar]]

[[Spike Prime - Arrows - Unax Murua, Eric Ayestaran, Leire Sáez de Cortázar |Arrows - Unax Murua, Eric Ayestaran, Leire Sáez de Cortázar]]

[[Gesture Controlled Tello Drone - Rapolas Kairys]]

[[Camera Controlled Robot Arm - Adrian Bartoš]]

[[Lego Spike Prime - Klára Senderáková, Štefan Beluško]]

[[Niryo Ned One - Ján Radovan Zubo, Martin Vitan, Gabriel Kožuch]]

[[Niryo Ned One - Townend Joseph]]

[[Lego Spike Prime Shooting game - Asier Mayoz, Ana Alonso]]

[[Lego Spike Prime Fishing game - Everest Alonso]]

[[RPI Step Motor controller - Viktor Mihalovic]]

== 2023 ==
[[Robot NICO IR senzor - Jakub Mišovský]]

[[Robot Eyes - Martin Slimák, Adam Struharňanský, Matej Magát]]

[[Janko Hraško - Miroslav Cibula, Tomáš Búcsi, Juraj Gavura]]

[[Jupiter-pohyb po fakulte - Tomáš Vikiszály]]

[[Kocúr v čižmách - Saša Snidová, Veronika Bendíková]]

[[Robot-sledovanie čiary/nasledovanie cieľa - Ondrej Adam, Tomáš Iványi]]

[[Čokoládová ruka - Martin Demovič]]

[[Dron sledovanie čiary -Patrik Homola]]

[[Spike Prime - Battleships - Teodor Fuček|Battleships - Teodor Fuček]]

[[Spike Prime - ColorSelect - Ákos Czére|ColorSelect - Ákos Czére]]

[[Kocúr v čižmách - Viktória Terebová]]

[[Niryo Ned One - Bálint Tatai]]

[[Pathfinder - Hana Hornáčková, Jozef Ševčík, Jelena Epifanic, Dario Lamaj]]

== 2021 ==

[[Robot-controlling Glove - Elisabet Delgado Mas]]

[[Kocur v Cizmach - Aljaž Andolšek and Ursek Slivsek]]

[[Ottis - Aljaž Andolšek and Ursek Slivsek]]

[[LunarLander (Deep) Q-learning - Filip Pavlove]]

[[Tello dron s OCR ovladanim - Jan Spirka]]

== 2020 ==

[[Pulzný oximeter a CO meter]]

[[Robot Otto očami začiatočníka - Patrik Modrovský]]

== 2019 ==

[[BOB808 - Peter Šeban, Jerguš Greššák]]

[[Tatrabot learns]]

[[Lilli object interaction - Michal Fikar, Nina Lúčna]]

[[MazeWalker - Beatriz Ramos, João Estorninho]]

[[100(1+1) - Lauren Kondratiev, Danijela Topić Vizcaya]]

== 2018 ==

[[Ottov plynulý pohyb a úprava prehrávania skladieb pomocou buzzera - Denis, Jana]]

[[Tatrabot 2D mapping - Johanna, Romana, Nicole, Oswaldo, Matej, Viktor]]

== 2017 ==

[[Tatrabot mapuje miestnost - Igor Slovak]]

[[BugBox(použitie DQN pri učení štvornohého robota optimálnej chôdze) - Martin Zvara]]

[[Identifikácia robotov na zákalde špecifických objektov pomocou webkamery - Dominik Knechta]]

[[Myš sledujúca čiaru - Marián Margeta a Marián Jonis]]

[[LCS in context of ML- Martin Mihál]]

[[ROS - Andrej Zemanovič]]

[[Robot ovládaný pomocou natrénovanej neurónovej siete - Katarína Kotrlová]]

[[POMDP - Drahomír Mrózek]]

== 2016 ==
[[V-Rep simulation|Space orientation robot in V-Rep - Dominik Kotvan a Ladislav Wagner]]

[[Tatrabot coordination|Tatrabot coordination - Ákos Hervay]]

[[Kladky s EV3|Kladky s EV3 - Michal Rakovský a Martin Palka]]

[[Tatrabot LineFollowing|Genetically evolved Neural Network for Tatrabot LineFollowing- Timotej Jurášek]]

[[Simulacia POMDP|Simulation of two-state Partially Observable Markov Decision Process - Roman Brojo]]

[[Monte carlo locatization with robot Mikes - Tomas Kubla]]

== 2015 ==
[[GolfBot|GolfBot - Andrej Jursa a Michal Moravčík]]

[[Sokoban|Sokoban - Dušan Suja, Máté Tibor Krajczár, Tamás Danis]]

== 2014 ==
[[Lego Mindstorms Ev3 Project - Ondrej Slovak]]

[[robot performs movements, detects trajectory using gyroscope/accellerometer, sends data to Java application that visualizes the trajectory (Mário Raček, Roman Moravec)]]

[[LEGO NXT project (Martin Marko, Martin Šalka)]]

[[GAZEBO simulator (Ľubomír Chriašteľ)]]

[[ASUS Xtion pre robota SmelyZajko (Marek Jelen, Miroslav Garai)]]

[[Robot Following using Camera (Marian Vysluzil)]]

== 2013 ==

[[Robot_constructs_a_map_of_a_maze_-_Adam_Bilisics]]

[[Šifrovačka s robotnačkou pre Euroveu, EV3 Puppy (Martina Chraščová, Lucia Budinská, Lucia Ďurikovičová )]]

[[E-Puck Bludisko]]

[[Visual navigation for Acrob or SBOT robot with Android with camera ( Marek Kádek, Jana Sucháneková)]]

[[Acrob robot constructs a map of a maze (Péter Dobsa, Gergely Labanc)]]

[[Acrob robot constructs a map of a maze - obstacles around the robot (Jozef Belko)]]

[[Trajectory from gyro/accellerometer/compas]]

[[Parking of robots in Remotely-operated laboratory]]

[[Pohyb robota po čiare a zaznamenávanie vzdialenosti okolitých objektov (Dudík, Kemény)]]

[[Agent Space on Raspberry Pi and Acrob with line-following and obstacle avoiding behavior-based robot.(Anton Blahunka, Martin Reiberger)]]

[[Evolution with Open Dynamics Engine (Bečvarová Ľuboslava)]]

[[Acrob navigation in maze using neural network (Kundlová, Jurenka)]]

[[ Acrob in maze (line sensor, distance sensor, robot navigates towards the maze exit) (Martin Hrivnák, Tomáš Matúš)]]

[[Acrob robot prejde bludisko, a program ďalej vizualizuje mapu bludiska na obrazovke (Michal Borčin & Barbora Zaťková)]]

[[NXT/LeJOs bot navigates maze consisting of black tape on white paper, object in front of robot is the maze exit (Pavol Lescinsky)]]

[[Omnibot(Krajči,Dluhý)]]

[[Navigation of the robot arm with ANN(Karin Vališová)]]

[[Evolving obstacle avoidance in Microsoft Robotics Studio (autori: Miroslav Matušťák & Pavol Kunovský)]]

[[Trajectory from gyro/accellerometer/compas with 9DOF IMU (Michal Zemko, Peter Svitok)]]

== 2012 ==

[[Avoiding in a corridor]]

[[Ketchup Robot]]

[[Identification of beer bottle using OpenCV]]

[[Mapping with SBOT]]

[[Light Avoiding SBOT]]

[[V-REP]]

[[Webots]]

[[Feedforward Backpropagation Neural Networks]]

[[AnyKode Marilou]]

== 2011 ==

[[Braitenberg 2]]

[[Tetrix robot with localisation using a neural network]]

[[Robot nasleduje čiaru a vykreslí svoju dráhu do programu]]

[[Mouse in Maze with robot E-Puck]]

[[Natural language interface for SBOT]]

[[SBOT following the line]]

[[Lego RCX following the line and obstacle avoiding ]]

[[Implement a simple LCS for the SBOT robot]]

[[Micromouse with Mindstorms]]

[[Mail delivery robot]]

[[Pohyb podla moznych krokov]]

[[New applet control interface for the lab control page ]]

[[Reproducing a learned geometrical object using drawing pen of Robotnacka robot]]

[[Obstacle avoidance training in Khepera 2 simulator via neural network]]

[[Highlander]]

== 2010 ==

[[Line Following Sbot Using Reinforcement Learning]]

[[Controlling robots in remotely-operated laboratory using Objection language]]

[[Robot Parking using NEAT - Neuroevolution of augmenting topologies]]

[[Gripper Functionality for Remotely-Operated Robotics Laboratory]]

[[Visualization of Robotnacka Trajectories in Robotics Laboratory]]

[[Robot whitch follow a line and stop back up of barrier]]

[[Robot with ultrasonic distance sensors]]

[[Braitenberg vehicles implemented using SBOT robot]]

[[Simple reactive Sbot agent]]

[[Simple python GUI for SBOT without packets]]

[[Controlling robot with webcam]]

[[SBOT pushing a vehicle]]

__notoc__

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:11:53Z

Robot:

File:Perp door zoom1 noted2.png

2025-07-02T21:11:36Z

Robot:

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:11:32Z

Robot:

= Robotická stavebnica pre Robotickú ligu – Sebastián Horváth =

== Cieľ projektu ==

Cieľom tohto projektu bolo navrhnúť základy prototypu '''robotickej stavebnice''' pre novú kategóriu v '''Robotickej Lige'''.

Táto stavebnica slúži ako '''didaktická pomôcka''', ktorá podporuje rozvoj detskej predstavivosti a súťaživosti v oblasti robotiky.

Inšpiráciou boli existujúce stavebnice ako '''LEGO Mindstorms''', '''Totem''' a '''Merkur'''.

Základom návrhu sú univerzálne stavebné "kocky", ktorých rozmery určujú kompatibilitu všetkých modulov.

== Realizácia projektu ==

=== Vlastnosti stavebnice ===

* '''Modularita''' – stavebnica je prispôsobiteľná rôznym typom úloh
* '''Jednoduchosť''' – vhodná na jednoduché použitie aj pre začiatočníkov
* '''3D tlačiteľnosť''' – všetky komponenty sú navrhnuté na domácu 3D tlač

=== Moduly ===

Vybrané prototypové moduly:

* Základné kocky
* Motor
* Joystick
* Senzor vzdialenosti

== Prototypy ==

=== Základné kocky ===

Základné stavebné prvky majú rozmery založené na štandarde skrutiek M3, s osovou vzdialenosťou 8 mm.
Dizajnovo pripomínajú diely zo stavebnice Merkur.

Kocky sú plne '''3D tlačiteľné''' a na konci dokumentu je uvedený odkaz na všetky modely.

[[File:Zaklad.png|250px]]
[[File:L.png|250px]]

=== Motor ===

Krabička pre servo motor bola navrhnutá tak, aby bola:

* kompaktná
* jednoducho tlačiteľná
* ergonomická

Servo sa vkladá zo spodnej strany krabičky a upevňuje pomocou skrutiek.
Ozubené koleso je navrhnuté tak, aby sa nepretáčalo a mohlo byť v prípade potreby aj priskrutkované.

'''Hlavné vlastnosti:'''
* Možnosť vytlačiť krabičku ako jeden celok
* Pevné uchytenie osi
* Jednoduchá výmena motora v prípade poruchy

[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]

=== Joystick ===

Prototyp umožňuje pevné pripojenie ku "kockám".
Na bokoch sú otvory pre matice M3, ktoré sa nepretáčajú vďaka presnému výrezu.
Vnútorné výstupky zabezpečujú stabilné uloženie joysticku.

'''Hlavné vlastnosti:'''
* Jednoduchosť
* Rýchle osadenie joysticku do krabičky
* Jednoduché pripojenie ku základným kockám

[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]

=== Senzor vzdialenosti ===

Základný a nevyhnutný modul – ultrazvukový senzor vzdialenosti – je navrhnutý tak, aby:

* bol jednoducho integrovaný do stavebnice
* bol odolný voči nárazom
* mal presné úchyty pre senzor a výrezy pre prístup k pinom

'''Hlavné vlastnosti:'''
* Odolná konštrukcia
* Presné uchytenie senzora
* Otvory pre jednoduché pripojenie k stavebnici
* Prístup ku konektorom cez spodnú časť krabičky

[[File:Ultrazvuk final.jpg|250px]]
[[File:Ultrazvuk.png|250px]]

== Nedostatky a možné úpravy ==

=== Motor ===

* Chýba možnosť pripevnenia ku stavebnici
* Nedostatočne stabilná podstava

→ V ďalšej verzii plánujem rozšíriť podstavu a pridať upevňovacie body.

=== Joystick a senzor ===

* Slabé uchytenie vrchného krytu
* Joystick sa pri používaní hýbe

→ Riešenia:
* Nahradiť zasúvací kryt za zacvakávací s drážkami
* Joystick bude priskrutkovaný do krabičky

== Ďalšie možnosti do budúcnosti ==

* Vývoj riadiacej jednotky s čipom '''RP2040-Zero''', napájanou lítiovou batériou
* Jednotka bude riadiť komunikáciu všetkých modulov
* Možnosť rozšírenia o '''gyroskop''' pre sledovanie natočenia robota (napr. pre autíčko)

== Záver ==

Tento projekt položil pevné základy pre vývoj '''robotickej stavebnice pre Robotickú ligu'''.
Modulárnosť, 3D tlač a jednoduchosť robia zo stavebnice skvelý nástroj na učenie robotiky u detí.

Do budúcna plánujem:
* Dokončiť riadiacu jednotku
* Opraviť prototypy podľa identifikovaných nedostatkov
* Rozšíriť počet funkčných modulov

[[File:Komplet.jpg|450px]]

== Video ukážka ==

[[File:Servo video.mp4]]

__NOTOC__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:11:20Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted2.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:07:09Z

Robot: /* Source code */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

[[File:explore.py|Corridor Walker code]]

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:05:29Z

Robot: /* Záver */

= Cieľ projektu =

Cieľom môjho projektu bolo navrhnúť základy prototypu robotickej stevebnice pre novú kategóriu v Robotickej Lige.
<br>Táto stavebnica budé prínosná ako didaktická pomôcka pre rozvíjanie detskej predstavivosti a kompetetivnosti v oblasti robotiky. <br> Hlavná inšpirácia boli stavebnice ako Lego Mindstorms, Totem a Merkur. Stavebnica má základné stavebné "kostičky", od ktorých sa odvíjajú rozmere všetkých modulov stavebnice.

= Realizácia projektu =

== Vlastnosti stavebnice ==

* Modularita - aby stavebnica sa dokázala prestavať pre iný typ úlohy
* Jednoduchosť - na pochopenie a použitie
* Možnosť ju vytlačiť 3D tlačiarňoou

== Moduly ==
Moduli (zatiaľ prototypy) pre túto stavebnicu som vybral:
*Základné kocky
*Motor
*Joystick
*Senzor na diaľku

== Prototypy ==

=== Základné kocky ===
Hlavné rozmery základných kociek boli velkosti dier pre skrutky M3, osová vzdialenosť 8mm a vyzerali ako základné diely v stavebnici Merkur.<br>
Tieto "kocky" sa dajú jednoducho vytlačiť na 3D tlačiarni a na konci dokumenu prikladám printables stránku so vsetkými 3D modelmi. <br>

<div>
[[File:Zaklad.png|250px]]
[[File:L.png|250px]]
</div>

=== Motor ===
Krabička (prototyp modulu) je jednoducho vytlačiteľná, kompaktná a ergonomická. <br>
Servo motorček sa dá vložiť zo spodu krabićky a následne zašróbované pomocou pridelenými šróbmi ku servo motorčeku. <br>
Ozubené kolo je spravené tak aby sa nepretáčalo pri pripevnení na rameno serva (je tu možnosť aj ozubené koleso priskrutkovať).

Hlavnou vlastnosťou motoru čo som chcel dovŕšiť boli:
*možnosť vytlačiť ako jeden celok,
*aby sa pevne otáćala os otáčania
*aby bolo jednoduché vymeniť motorček pri jeho prípadnom zhoraní.

<div>
[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]
</div>
=== Joystick ===
Tento prototyp je jeden z prvých čo má možnosť pripevnenie sa ku "kockám" a prepojiť tak ku celej stavebnici.<br>
Diery na bokoch sluzia pre matice do ktorych sa zaskrutkuje kocka. Matice sa nebude pretacat, kedze diera je priamo na velkost M3 matice.<br>
V nutri su vystupky presne pre uchytenie joysticku.

Hlavné vlastnosti:
*Jednoduchosť
*Ľahké vloženie joysticku do krabičky
*Jednoducho pripojiťeľné ku "kockám"

<div>
[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]
</div>

=== Senzor ===
Jednoduchý ale nevyhnutny komponent v stavebnici je senzor vzdialenosti. Pre zakladne uceli nasej stavebnice chcem aby sa dal prepojit ku celej stavebnici.<br>
To sa umoznilo pomocou dier ako pri joysticku. Dalej jeho pevnost, krabicka je dostatocne odolna proti narazom. Pre pinout sluzi aj diera pre piny na spodku krabicky. <br>
Taktiez ma aj vystupky pre uchytenie samotneho ultrazvukoveho senzora.

Hlavne vlastnosti:
//todo dopln podla opysu vyssie
<div>
[[File:Ultrazvuk final.jpg|250px]]
[[File:Ultrazvuk.png|250px]]
</div>

== Chyby a mozna uprava ==
=== Motor ===
Nema pripevnenia ku stavebnici a nema dobru stabilnu polohu na ktorej by mohola krabicka stat. Tieto chyby chcem prerobit v dalsich prototypoch pridanim pripevneni a zvacsenia podstavy.

=== Joystick + Senzor ===
Pripevnenie poklicky na krabicke je slabe a jednoducho sa odpaja, joystick zle sedi na panloch a hybe sa.<br>
Ponukaju sa jednoduche riesenia: namiesto zasuvania sa do panelov, tak sa joystick priskrutkuje do krabicky.<br>
Dalej jednoduchuj poklicku co sa "len tak" pripevni ku celku vymenim za poklop so zubami na zacvaknutie.

== Ďalšie moznosti do buducna==
Pre rozsirenie prototypu stavebnice chcem spravit hlavnu riadiacu jednotku ktora bude napjena na litiove baterky a bude riadi celu stavebnicu. <br>
Ta bude z plosneho spoju Raspebby RP2040-zero, ktory bez problemu bude riadit celu stavebnicu.<br>
<br>
Dalej je urcite dobry napad zakomponovat gyroskop ktory bude zaznamenavat natocenie robota (pripadne auticko).

= Záver =
Na zaver mam dobry zaklad na roboticku stavebnicu pre Roboticku ligu, z ktorej sa deti budu moct ucit. //todo dopln

<div>
[[File:Komplet.jpg|450px]]
</div>

= Video =
[[File:Servo video.mp4]]

__notoc__

File:Komplet.jpg

2025-07-02T21:05:03Z

Robot:

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:04:16Z

Robot: /* Záver */

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T21:03:38Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:03:37Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|1000px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:02:58Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|800px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:02:42Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|600px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:02:29Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|500px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|500px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Glass doors.png

2025-07-02T21:01:44Z

Robot:

File:Glass doors life.jpg

2025-07-02T21:01:32Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T21:01:17Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|400px]]

4. Gap detection and odometry integration.

[[Image:glass_doors_life.jpg|500px]]
[[Image:glass_doors.png|400px]]

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Before corridor.png

2025-07-02T20:57:47Z

Robot:

File:Corridor life.jpg

2025-07-02T20:57:33Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:57:17Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.
[[Image:corridor_life.jpg|500px]]
[[Image:before_corridor.png|400px]]

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:55:43Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|500px]]

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Artefact rods.png

2025-07-02T20:55:15Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:55:03Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]
[[Image:artefact_rods.png|400px]]

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:53:00Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|400px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|400px]]

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Rods lidar r.png

2025-07-02T20:52:31Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:52:17Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|300px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar_r.png|300px]]

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Rods lidar.png

2025-07-02T20:49:24Z

Robot:

File:Robot construction.jpg

2025-07-02T20:48:56Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:48:36Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|300px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

[[Image:robot_construction.jpg|500px]]
[[Image:rods_lidar.png|300px]]

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T20:42:07Z

Robot: /* Senzor */

= Cieľ projektu =

Cieľom môjho projektu bolo navrhnúť základy prototypu robotickej stevebnice pre novú kategóriu v Robotickej Lige.
<br>Táto stavebnica budé prínosná ako didaktická pomôcka pre rozvíjanie detskej predstavivosti a kompetetivnosti v oblasti robotiky. <br> Hlavná inšpirácia boli stavebnice ako Lego Mindstorms, Totem a Merkur. Stavebnica má základné stavebné "kostičky", od ktorých sa odvíjajú rozmere všetkých modulov stavebnice.

= Realizácia projektu =

== Vlastnosti stavebnice ==

* Modularita - aby stavebnica sa dokázala prestavať pre iný typ úlohy
* Jednoduchosť - na pochopenie a použitie
* Možnosť ju vytlačiť 3D tlačiarňoou

== Moduly ==
Moduli (zatiaľ prototypy) pre túto stavebnicu som vybral:
*Základné kocky
*Motor
*Joystick
*Senzor na diaľku

== Prototypy ==

=== Základné kocky ===
Hlavné rozmery základných kociek boli velkosti dier pre skrutky M3, osová vzdialenosť 8mm a vyzerali ako základné diely v stavebnici Merkur.<br>
Tieto "kocky" sa dajú jednoducho vytlačiť na 3D tlačiarni a na konci dokumenu prikladám printables stránku so vsetkými 3D modelmi. <br>

<div>
[[File:Zaklad.png|250px]]
[[File:L.png|250px]]
</div>

=== Motor ===
Krabička (prototyp modulu) je jednoducho vytlačiteľná, kompaktná a ergonomická. <br>
Servo motorček sa dá vložiť zo spodu krabićky a následne zašróbované pomocou pridelenými šróbmi ku servo motorčeku. <br>
Ozubené kolo je spravené tak aby sa nepretáčalo pri pripevnení na rameno serva (je tu možnosť aj ozubené koleso priskrutkovať).

Hlavnou vlastnosťou motoru čo som chcel dovŕšiť boli:
*možnosť vytlačiť ako jeden celok,
*aby sa pevne otáćala os otáčania
*aby bolo jednoduché vymeniť motorček pri jeho prípadnom zhoraní.

<div>
[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]
</div>
=== Joystick ===
Tento prototyp je jeden z prvých čo má možnosť pripevnenie sa ku "kockám" a prepojiť tak ku celej stavebnici.<br>
Diery na bokoch sluzia pre matice do ktorych sa zaskrutkuje kocka. Matice sa nebude pretacat, kedze diera je priamo na velkost M3 matice.<br>
V nutri su vystupky presne pre uchytenie joysticku.

Hlavné vlastnosti:
*Jednoduchosť
*Ľahké vloženie joysticku do krabičky
*Jednoducho pripojiťeľné ku "kockám"

<div>
[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]
</div>

=== Senzor ===
Jednoduchý ale nevyhnutny komponent v stavebnici je senzor vzdialenosti. Pre zakladne uceli nasej stavebnice chcem aby sa dal prepojit ku celej stavebnici.<br>
To sa umoznilo pomocou dier ako pri joysticku. Dalej jeho pevnost, krabicka je dostatocne odolna proti narazom. Pre pinout sluzi aj diera pre piny na spodku krabicky. <br>
Taktiez ma aj vystupky pre uchytenie samotneho ultrazvukoveho senzora.

Hlavne vlastnosti:
//todo dopln podla opysu vyssie
<div>
[[File:Ultrazvuk final.jpg|250px]]
[[File:Ultrazvuk.png|250px]]
</div>

== Chyby ==

== Ďalšie úpravy ==

= Záver =

V závere mám úplne funkčnú robotickú ruku, ktorá dokáže namáčať jahodu do čokolády, dostatočne ju odkvapkať rôznymi pohybmi špecifickými pre hustú čokoládu a špecifiká čerstvého namáčaného ovocia a vrátiť ju naspäť na pôvodné miesto. Vo vnútri kódu som nechal komentáre, aby bolo ľahké orientovať sa v jednotlivých častiach.

= Ďalšie nápady =
* Pohyblivá sieťka v dvoch osiach x a y - na držanie napichnutého ovocia s väčším počtom pozícii, ktorá robotovi umožní spracovať samostatne kompletnú sadu ovocia. <br>
* Sekcia pre posypanie jahôd posýpkami (napr. jemne namleté orechy) <br>
* Naprogramovať ruku, aby vedela vyschnuté ovocie doplnkovo očokoládovať drobnými čiarkami čokolády kontrastnej farby (biela, čierna), čiže vytvoriť ornamenty na jahodách. <br>

= Fotografie a popisy =
Ruka po úpravách
<div>
[[Image:Ruka.jpeg|border|400px]]
</div>

<div>
CCPM Servo Consistency Master (servo-tester) – Veľmi užitočná pomôcka s ktorou som sa vedel uistiť, že moje servo-motory sú na neutrály a skontrolovať si range of motion.

[[Image:Neutral.jpeg|border|330px]]
</div>

= Video =
<youtube width="720" height="396">mIBqqe55oyY</youtube><br>

= Source Code =
[[Media:RoboArm.zip|RoboArm.zip]]

__notoc__

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T20:41:42Z

Robot: /* Prototypy */

= Cieľ projektu =

Cieľom môjho projektu bolo navrhnúť základy prototypu robotickej stevebnice pre novú kategóriu v Robotickej Lige.
<br>Táto stavebnica budé prínosná ako didaktická pomôcka pre rozvíjanie detskej predstavivosti a kompetetivnosti v oblasti robotiky. <br> Hlavná inšpirácia boli stavebnice ako Lego Mindstorms, Totem a Merkur. Stavebnica má základné stavebné "kostičky", od ktorých sa odvíjajú rozmere všetkých modulov stavebnice.

= Realizácia projektu =

== Vlastnosti stavebnice ==

* Modularita - aby stavebnica sa dokázala prestavať pre iný typ úlohy
* Jednoduchosť - na pochopenie a použitie
* Možnosť ju vytlačiť 3D tlačiarňoou

== Moduly ==
Moduli (zatiaľ prototypy) pre túto stavebnicu som vybral:
*Základné kocky
*Motor
*Joystick
*Senzor na diaľku

== Prototypy ==

=== Základné kocky ===
Hlavné rozmery základných kociek boli velkosti dier pre skrutky M3, osová vzdialenosť 8mm a vyzerali ako základné diely v stavebnici Merkur.<br>
Tieto "kocky" sa dajú jednoducho vytlačiť na 3D tlačiarni a na konci dokumenu prikladám printables stránku so vsetkými 3D modelmi. <br>

<div>
[[File:Zaklad.png|250px]]
[[File:L.png|250px]]
</div>

=== Motor ===
Krabička (prototyp modulu) je jednoducho vytlačiteľná, kompaktná a ergonomická. <br>
Servo motorček sa dá vložiť zo spodu krabićky a následne zašróbované pomocou pridelenými šróbmi ku servo motorčeku. <br>
Ozubené kolo je spravené tak aby sa nepretáčalo pri pripevnení na rameno serva (je tu možnosť aj ozubené koleso priskrutkovať).

Hlavnou vlastnosťou motoru čo som chcel dovŕšiť boli:
*možnosť vytlačiť ako jeden celok,
*aby sa pevne otáćala os otáčania
*aby bolo jednoduché vymeniť motorček pri jeho prípadnom zhoraní.

<div>
[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]
</div>
=== Joystick ===
Tento prototyp je jeden z prvých čo má možnosť pripevnenie sa ku "kockám" a prepojiť tak ku celej stavebnici.<br>
Diery na bokoch sluzia pre matice do ktorych sa zaskrutkuje kocka. Matice sa nebude pretacat, kedze diera je priamo na velkost M3 matice.<br>
V nutri su vystupky presne pre uchytenie joysticku.

Hlavné vlastnosti:
*Jednoduchosť
*Ľahké vloženie joysticku do krabičky
*Jednoducho pripojiťeľné ku "kockám"

<div>
[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]
</div>

=== Senzor ===
Jednoduchý ale nevyhnutny komponent v stavebnici je senzor vzdialenosti. Pre zakladne uceli nasej stavebnice chcem aby sa dal prepojit ku celej stavebnici.<br>
To sa umoznilo pomocou dier ako pri joysticku. Dalej jeho pevnost, krabicka je dostatocne odolna proti narazom. Pre pinout sluzi aj diera pre piny na spodku krabicky. <br>
Taktiez ma aj vystupky pre uchytenie samotneho ultrazvukoveho senzora.

Hlavne vlastnosti:
//todo dopln podla opysu vyssie

[[File:Ultrazvuk.png|250px]]
[[File:Ultrazvuk final.jpg|250px]]

== Chyby ==

== Ďalšie úpravy ==

= Záver =

V závere mám úplne funkčnú robotickú ruku, ktorá dokáže namáčať jahodu do čokolády, dostatočne ju odkvapkať rôznymi pohybmi špecifickými pre hustú čokoládu a špecifiká čerstvého namáčaného ovocia a vrátiť ju naspäť na pôvodné miesto. Vo vnútri kódu som nechal komentáre, aby bolo ľahké orientovať sa v jednotlivých častiach.

= Ďalšie nápady =
* Pohyblivá sieťka v dvoch osiach x a y - na držanie napichnutého ovocia s väčším počtom pozícii, ktorá robotovi umožní spracovať samostatne kompletnú sadu ovocia. <br>
* Sekcia pre posypanie jahôd posýpkami (napr. jemne namleté orechy) <br>
* Naprogramovať ruku, aby vedela vyschnuté ovocie doplnkovo očokoládovať drobnými čiarkami čokolády kontrastnej farby (biela, čierna), čiže vytvoriť ornamenty na jahodách. <br>

= Fotografie a popisy =
Ruka po úpravách
<div>
[[Image:Ruka.jpeg|border|400px]]
</div>

<div>
CCPM Servo Consistency Master (servo-tester) – Veľmi užitočná pomôcka s ktorou som sa vedel uistiť, že moje servo-motory sú na neutrály a skontrolovať si range of motion.

[[Image:Neutral.jpeg|border|330px]]
</div>

= Video =
<youtube width="720" height="396">mIBqqe55oyY</youtube><br>

= Source Code =
[[Media:RoboArm.zip|RoboArm.zip]]

__notoc__

File:Perp door zoom1 noted.png

2025-07-02T20:34:30Z

Robot:

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:34:19Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1_noted.png|300px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T20:31:24Z

Robot: /* Joystick */

= Cieľ projektu =

Cieľom môjho projektu bolo navrhnúť základy prototypu robotickej stevebnice pre novú kategóriu v Robotickej Lige.
<br>Táto stavebnica budé prínosná ako didaktická pomôcka pre rozvíjanie detskej predstavivosti a kompetetivnosti v oblasti robotiky. <br> Hlavná inšpirácia boli stavebnice ako Lego Mindstorms, Totem a Merkur. Stavebnica má základné stavebné "kostičky", od ktorých sa odvíjajú rozmere všetkých modulov stavebnice.

= Realizácia projektu =

== Vlastnosti stavebnice ==

* Modularita - aby stavebnica sa dokázala prestavať pre iný typ úlohy
* Jednoduchosť - na pochopenie a použitie
* Možnosť ju vytlačiť 3D tlačiarňoou

== Moduly ==
Moduli (zatiaľ prototypy) pre túto stavebnicu som vybral:
*Základné kocky
*Motor
*Joystick
*Senzor na diaľku

== Prototypy ==

=== Základné kocky ===
Hlavné rozmery základných kociek boli velkosti dier pre skrutky M3, osová vzdialenosť 8mm a vyzerali ako základné diely v stavebnici Merkur.

Tieto "kocky" sa dajú jednoducho vytlačiť na 3D tlačiarni a na konci dokumenu prikladám printables stránku so vsetkými 3D modelmi.

<div>
[[File:Zaklad.png|250px]]
[[File:L.png|250px]]
</div>

=== Motor ===
Hlavnou vlastnosťou motoru čo som chcel dovŕšiť boli:
*možnosť vytlačiť ako jeden celok,
*aby sa pevne otáćala os otáčania
*aby bolo jednoduché vymeniť motorček pri jeho prípadnom zhoraní.

Krabička (prototyp modulu) je jednoducho vytlačiteľná, kompaktná a ergonomická. <br>
Servo motorček sa dá vložiť zo spodu krabićky a následne zašróbované pomocou pridelenými šróbmi ku servo motorčeku. <br>
Ozubené kolo je spravené tak aby sa nepretáčalo pri pripevnení na rameno serva (je tu možnosť aj ozubené koleso priskrutkovať).

[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]

=== Joystick ===
Hlavné vlastnosti:
*Jednoduchosť
*Ľahké vloženie joysticku do krabičky
*Jednoducho pripojiťeľné ku "kockám"

Tento prototyp je jeden z prvých čo má možnosť pripevnenie sa ku "kockám" a prepojiť tak ku celej stavebnici.<br>
Diery na bokoch sluzia pre matice do ktorych sa zaskrutkuje kocka. Matice sa nebude pretacat, kedze diera je priamo na velkost M3 matice.<br>
V nutri su vystupky presne pre uchytenie joysticku.

[[File:Joystick final.jpg|250px]]
[[File:Joystick krabicka.png|250px]]

=== Senzor ===

== Chyby ==

== Ďalšie úpravy ==

= Záver =

V závere mám úplne funkčnú robotickú ruku, ktorá dokáže namáčať jahodu do čokolády, dostatočne ju odkvapkať rôznymi pohybmi špecifickými pre hustú čokoládu a špecifiká čerstvého namáčaného ovocia a vrátiť ju naspäť na pôvodné miesto. Vo vnútri kódu som nechal komentáre, aby bolo ľahké orientovať sa v jednotlivých častiach.

= Ďalšie nápady =
* Pohyblivá sieťka v dvoch osiach x a y - na držanie napichnutého ovocia s väčším počtom pozícii, ktorá robotovi umožní spracovať samostatne kompletnú sadu ovocia. <br>
* Sekcia pre posypanie jahôd posýpkami (napr. jemne namleté orechy) <br>
* Naprogramovať ruku, aby vedela vyschnuté ovocie doplnkovo očokoládovať drobnými čiarkami čokolády kontrastnej farby (biela, čierna), čiže vytvoriť ornamenty na jahodách. <br>

= Fotografie a popisy =
Ruka po úpravách
<div>
[[Image:Ruka.jpeg|border|400px]]
</div>

<div>
CCPM Servo Consistency Master (servo-tester) – Veľmi užitočná pomôcka s ktorou som sa vedel uistiť, že moje servo-motory sú na neutrály a skontrolovať si range of motion.

[[Image:Neutral.jpeg|border|330px]]
</div>

= Video =
<youtube width="720" height="396">mIBqqe55oyY</youtube><br>

= Source Code =
[[Media:RoboArm.zip|RoboArm.zip]]

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:31:21Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1.png|300px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__

File:Perp door zoom1.png

2025-07-02T20:30:55Z

Robot:

Robotická stavebnica pre Robotickú ligu - Sebastián Horváth

2025-07-02T20:30:52Z

Robot: /* Joystick */

= Cieľ projektu =

Cieľom môjho projektu bolo navrhnúť základy prototypu robotickej stevebnice pre novú kategóriu v Robotickej Lige.
<br>Táto stavebnica budé prínosná ako didaktická pomôcka pre rozvíjanie detskej predstavivosti a kompetetivnosti v oblasti robotiky. <br> Hlavná inšpirácia boli stavebnice ako Lego Mindstorms, Totem a Merkur. Stavebnica má základné stavebné "kostičky", od ktorých sa odvíjajú rozmere všetkých modulov stavebnice.

= Realizácia projektu =

== Vlastnosti stavebnice ==

* Modularita - aby stavebnica sa dokázala prestavať pre iný typ úlohy
* Jednoduchosť - na pochopenie a použitie
* Možnosť ju vytlačiť 3D tlačiarňoou

== Moduly ==
Moduli (zatiaľ prototypy) pre túto stavebnicu som vybral:
*Základné kocky
*Motor
*Joystick
*Senzor na diaľku

== Prototypy ==

=== Základné kocky ===
Hlavné rozmery základných kociek boli velkosti dier pre skrutky M3, osová vzdialenosť 8mm a vyzerali ako základné diely v stavebnici Merkur.

Tieto "kocky" sa dajú jednoducho vytlačiť na 3D tlačiarni a na konci dokumenu prikladám printables stránku so vsetkými 3D modelmi.

<div>
[[File:Zaklad.png|250px]]
[[File:L.png|250px]]
</div>

=== Motor ===
Hlavnou vlastnosťou motoru čo som chcel dovŕšiť boli:
*možnosť vytlačiť ako jeden celok,
*aby sa pevne otáćala os otáčania
*aby bolo jednoduché vymeniť motorček pri jeho prípadnom zhoraní.

Krabička (prototyp modulu) je jednoducho vytlačiteľná, kompaktná a ergonomická. <br>
Servo motorček sa dá vložiť zo spodu krabićky a následne zašróbované pomocou pridelenými šróbmi ku servo motorčeku. <br>
Ozubené kolo je spravené tak aby sa nepretáčalo pri pripevnení na rameno serva (je tu možnosť aj ozubené koleso priskrutkovať).

[[File:Servo_final.jpg|250px]]
[[File:Servo2.png|250px]]

=== Joystick ===
Hlavné vlastnosti:
*Jednoduchosť
*Ľahké vloženie joysticku do krabičky
*Jednoducho pripojiťeľné ku "kockám"

Tento prototyp je jeden z prvých čo má možnosť pripevnenie sa ku "kockám" a prepojiť tak ku celej stavebnici.<br>
Diery na bokoch sluzia pre matice do ktorych sa zaskrutkuje kocka. Matice sa nebude pretacat, kedze diera je priamo na velkost M3 matice.<br>
V nutri su vystupky presne pre uchytenie joysticku.

[[File:Joystick final.png]]
[[File:Joystick krabicka.png]]

=== Senzor ===

== Chyby ==

== Ďalšie úpravy ==

= Záver =

V závere mám úplne funkčnú robotickú ruku, ktorá dokáže namáčať jahodu do čokolády, dostatočne ju odkvapkať rôznymi pohybmi špecifickými pre hustú čokoládu a špecifiká čerstvého namáčaného ovocia a vrátiť ju naspäť na pôvodné miesto. Vo vnútri kódu som nechal komentáre, aby bolo ľahké orientovať sa v jednotlivých častiach.

= Ďalšie nápady =
* Pohyblivá sieťka v dvoch osiach x a y - na držanie napichnutého ovocia s väčším počtom pozícii, ktorá robotovi umožní spracovať samostatne kompletnú sadu ovocia. <br>
* Sekcia pre posypanie jahôd posýpkami (napr. jemne namleté orechy) <br>
* Naprogramovať ruku, aby vedela vyschnuté ovocie doplnkovo očokoládovať drobnými čiarkami čokolády kontrastnej farby (biela, čierna), čiže vytvoriť ornamenty na jahodách. <br>

= Fotografie a popisy =
Ruka po úpravách
<div>
[[Image:Ruka.jpeg|border|400px]]
</div>

<div>
CCPM Servo Consistency Master (servo-tester) – Veľmi užitočná pomôcka s ktorou som sa vedel uistiť, že moje servo-motory sú na neutrály a skontrolovať si range of motion.

[[Image:Neutral.jpeg|border|330px]]
</div>

= Video =
<youtube width="720" height="396">mIBqqe55oyY</youtube><br>

= Source Code =
[[Media:RoboArm.zip|RoboArm.zip]]

__notoc__

Towards Localization and Navigation in a Linear Corridor with Jupiter Robot - Anastasiya Ihnatovich

2025-07-02T20:30:12Z

Robot: /* Project's stages */

== Goals of the project ==

The goal of this project is to explore approaches for potential navigation using LiDAR data in the corridors of FMFI UK. The focus is on detecting and utilizing environmental features such as walls and doors for potential localization and movement.

== Project's stages ==

1. Defining the area of interest: localization using data from LiDAR.

2. Implementing straight movement: basic module for moving straight along a corridor.

3. Real-time visualization of LiDAR data along the movement to analyze potentially useful features during movement.

4. Analysing LiDAR's data. Following features were found:

* Glass doors create detectable gaps in LiDAR scans when the robot is not perpendicular to them.

[[Image:perp_door_life.jpg|500px]]
[[Image:perp_door_zoom1.png|500px]]

* Metal construction rods of the robot itself interfere with LiDAR beams, creating artificial gaps.
Hypothesis: The sensor requires adjustment time due to significant differences in distances between these rods and their surroundings. This causes distortion in scan data (picture corresponds the situation when LiDAR rotates clockwise).

* Corridors generally consist of nearly perpendicular to main walls segments.

4. Gap detection and odometry integration.

== Description of the project ==

The program consists of mainly parts:

1. Getting linear functions which represent walls using library RANSAC algorithm. The projects' limitations and exploration allowed to use the right wall as orientation of straight movement: if k coefficient (y=kx+b, edge case was eliminated) is more than 0, robot turns a little to the right, otherwise, to the left. Such approach showed good results in the given environment.

2. Finding consequent points close to walls, which have a distance more than "limit" (experiments showed that 0.7 m. gives good results in the given environment). These gaps represent doors, which potentially can be used to refine position, given by odometry, to navigate.

3. Visualization of Lidar data and found gaps in real-time.

== Video ==

videos of a system working

<youtube width="390" height="700">jz0L_NojaEI</youtube>
<youtube width="800" height="600">q3ujcNBji2U</youtube>

== Source code ==

link to github or uploaded file with code

== Resources ==

RANSAC algorithm: https://scikit-learn.org/stable/modules/linear_model.html#ransac-regression

__notoc__