Teaching
Lecturer in Robotics, Automation and Control Engineering at the University of Brighton, teaching across robotics, machine learning and programming.
“I connect robotics and AI theory with programming, simulation and hands-on experimentation, helping students understand both the mathematical principles and their implementation in real systems.”
Approach
01
Every equation on a slide should have a counterpart students can execute — a kinematics derivation becomes a ROS 2 node, a loss function becomes a training curve they can break and fix.
02
Students build confidence in RViz, PyBullet or a dry-run mode first, then move to real equipment knowing what the robot should do and how to stop it.
03
Assessment is most useful when it tells a student what to try next. I favour short, specific feedback on code and design decisions over generic marks.
Current teaching
Teaching contributions in the School of Architecture, Technology and Engineering and in computing.
Robotics, Automation and Control Engineering
University of Brighton
Machine Learning
University of Brighton · Level 6
Introduction to Programming (CI401)
University of Brighton
Working in the Computing Industry (CI453)
University of Brighton
Enterprise Networking
University of Brighton
Cyber Security Management
University of Brighton
Previous teaching
Physics, Mathematics, Engineering and Programming
Kaplan International Pathways
Mechanical Engineering
Undergraduate teaching, India
In practice
Guiding undergraduate and postgraduate projects in robotics, computer vision and applied machine learning, from scoping to write-up.
Live demonstrations on the Franka Emika Panda and in simulation, so students see planning, control and safety behaviour in practice.
Short, incremental exercises in Python and ROS 2 that students extend week by week.
Coursework and practical assessment with written feedback focused on the next improvement.
Workshop topics
Teaching resources
Both packages target ROS 2 Humble on Ubuntu 22.04 and run entirely in simulation.
sh1755/ROS2-Panda-Pick-and-Place
Problem. Students need a complete, readable example of motion planning, planning scenes and gripper control.
Approach. A teaching package that adds a table and cube to the MoveIt planning scene, plans pose-constrained motions, commands the gripper and attaches the object through a full pick-and-place sequence in RViz.
Status: Teaching resource
sh1755/ROS2-Forward-Inverse-Kinematics-repository
Problem. FK and IK are usually taught on paper; students rarely see a redundant arm respond to their own maths.
Approach. FK by chained transforms and damped-least-squares Jacobian IK, published as joint states and animated in RViz 2 through approach, grasp, lift, transfer and place stages.
Status: Teaching resource