Reference no: EM133051720 , Length: word count:3000
Robotics
Learning outcome 1: Navigation control of a mobile robot, and the analysis and correct application of the Potential Field method to robot navigation
Learning outcome 2: Understand the basic parts and functionalities of a mobile robot
Learning outcome 3: Understand how to combine perception, action and planning in a real-world environment and demonstrate this in simulation
Mobile Robot Navigation Problem: Design and implementation.
Navigation in an environment that can be unstructured and partially unknown is a key issue for a mobile robot.
Navigation is closely related to robot motion planning that can be described as going to a target without hitting obstacles and other robots (and humans)
You are required to:
Part A
Description of a navigation problem
This should include a description of the environment, determination of how many robots act (navigate) in the environment (one, two or three), what types of sensors your robots use, what tasks the robots should perform and what navigation procedure is applied by the robot.
1. Environment (world, workspace)
Determine the environment where your robots will act. The environment should include obstacles (either round or polygonal) and their positions, a target and its position and the initial position of your robot.
If your navigation scenario involves more than one robot, there can be more than one target (e.g. one target for each robot).
You draw your environment by using MATLAB.
2. Parameters of a robot ( robots):
Robots can be considered as points with masses equal to one. However, it is also possible to consider a robot as a certain shape, such as round or rectangular.
a) What kind of locomotion does your robot use?
Is it a synchro-drive, omnidirectional or a differential drive method? (Chapter 1, Autonomous Mobile Robots, R. Siegwart et al)
b) What sensors is your robot equipped with?
In you scenario you can use only one (virtual) sensor, that can possess either limited or perfect (unlimited) visibility.
In general, there are always sensors that allow to 'sense' obstacles (infrared or sonar).
d) If your scenario includes more than one robot, is one robot a leader? Or the robots act independently?
e) Are your robots identical or they are different? For example, if your sensor is vision, each robot can have a different colour, and this may allow robots recognising other robots.
3. Tasks
a) What kind of tasks should the robots perform?
b) Each robot goes to its corresponding goal or do all robots have the same goal?
For example, each robot has garbage that should be disposed at a certain location. In this case, all robots may have the same target. However, if the robots act in a warehouse they can go different locations (e.g. to collect tools or equipment).
4. Navigation control
a) What control should be applied so robots do not collide with obstacles and each other?
b) What motor schemas should be used?
For example, go to Goal, Avoid obstacles, Avoid collisions with other robots?
c) How robots can find the goal?
d) Describe your robot's behaviour/motion; you may assume that your robot is omnidirectional and holonomic.
An option: a robot knows the coordinates of the GOAL, they are given. In the case the robots have perfect sensors; they can sense all obstacles in the environment. If their sensors are not perfect, robots can follow the direction to the goal, and when it encounters an obstacle, the direction to follow is updated.
Note. For navigation control use the artificial potential field approach. Goals are modelled as attraction potentials, and obstacles - as repulsive potentials.
Correspondingly, obstacles create repulsive forces towards robots. A robot creates either repulsive or attraction force towards other robots. The latter case can occur if a robot follows the leader. Goals create attractive forces. Robot movement is controlled by the resulting force which is the weighted sum of all forces acting on a robot. We can consider a simplified situation: velocity of a robot coincides with a) direction of the resulting force; b) magnitude of the resulting force with some scaling coefficient for a unit time step.
Describe possible conflict situations.
The description of all the steps, that robot need to perform to reach their targets, is the basis for the pseudocode. Samples of pseudocodes will be provided.
e) Providing a code for artificial potential function. You can attempt to design the potential function for your scenario; however you can provide a code of a simple potential function, consisting of one attraction and one repulsion potentials. Do not forget to comment your code.
Samples of codes that visualise potential functions, and their (negative) gradients will be given and designed during the tutorials. You can use the existing codes as a basis, but you need to update them by changing some parameters and indicate what changes have been made. You should also indicate your source.
Further guidance will be given throughout the classes and laboratory sessions on suggested network types to be investigated and the scope and depth required.
Part B
Case study: robots and their applications
Below is the list of potential application areas of robotics:
1. Manufacturing
2. Healthcare
3. Arts and Entertainment
4. Domestic use
5. Military and security
6. Search and rescue
7. Education
8. Care of children, elderly, and the disabled
9. Space exploration
10. Agriculture
11. Transport\ Logistics
12. Toys and Leisure
Please, choose any 3 (three) areas from the list. You should:
1) Justify your choice, e.g. why and how do you think that those areas can benefit from the use of robots. Address each areas separately.
2) Discuss and highlight the required characteristics and specifications which would be desired for a mobile robot operating in your chosen application areas.
Your answer should address attributes related to hardware, functionalities, aesthetics and human/social interaction. If possible, illustrate your answers.
Attachment:- Robotics.rar