¹û¶³Ó°Ôº

The unique characteristic which sets robots apart from other autonomous systems is that they can move themselves. A key capability is that the robot must plan how it will move through space to achieve its desired tasks.

This module covers two tightly interlinked topics: modelling and motion planning. Modelling is means by which a robot is mathematically described. It includes transformations, coordinate frames, forward and inverse kinematics and manipulability. Motion planning refers to specifying a sequence of transformations the robot must undertake to complete tasks. It is typically determined using graph search and sampling techniques. This module will expose students to basic theory, current leading-edge research, and the strengths and weaknesses of different approaches. This evaluation will include technical, ethical and societal impacts. It complements Estimation and Control which looks at how to carry out the low-level estimation and control of individual joints and actuators.

Aims:

The aims of this module are to:

• Develop students’ knowledge of how to mathematically describe a robot and the algorithms which can be used to specify the desired values that joints and actuators should take to achieve a task – theoretically how they work, how they are realised, and when is it most appropriate for the different types and kinds of control.
• Support students in creating practical solutions in robotics and AI against functional and non-functional requirements, testing and assessing those in simulated and real-world environments and articulating the limitations of those assessments.
• Provide the students with tools for critical analysis of reasoning about the appropriateness and quality of practical solutions produced in the context of the problems defined.

Intended learning outcomes:

On successful completion of the module, a student will be able to:

1. Mathematically model a robot in terms of actuators and joints, with the goal of applying this in the context of control to solve complex problems in the field of robotics and AI.
2. Apply fundamentals of motion planning, in the context of control to solve complex problems in the field of robotics and Artificial Intelligence.
3. Use tools such as differential equations, control theory, and simulation software to develop mathematical models of robotic systems and other dynamic systems, using tools such as differential equations, control theory, and simulation software.
4. Analyse and evaluate the performance of different motion planning algorithms based on relevant criteria, such as efficiency, accuracy, and stability.
5. Identify and implement possible solutions to address the limitations and challenges of motion planning, such as uncertainty, nonlinearity, and complexity.

Indicative content:

The following are indicative of the topics the module will typically cover:

• Coordinate frames.
• Forward and inverse kinematics.
• Manipulatibility (Jacobians)
• Graph-based search algorithms (A*, Dijkstra.)
• Rapidly exploring random trees.
• Kinodynamic planning.
• Smoothing algorithms including time elastic band.
• Energy efficiency.
• Ethics and societal impacts.

Requisite conditions:

To be eligible to select this module as optional or elective, a student must be registered on a programme and year of study for which it is formally available.