Constrained torque-level formation control via consensus FBL-MPC
Torque-level formation control for differential-drive robots. A consensus MPC on feedback-linearised dynamics, posed as a convex program that solves in real time.
Torque-level formation control for differential-drive robots. A consensus MPC on feedback-linearised dynamics, posed as a convex program that solves in real time.
Adaptive impedance control for a 7-DOF manipulator with formal safety guarantees under state and input limits, deployed on KINOVA and Franka arms.
Stability analysis and design of high-order, multi-constrained safety filters for linear systems.
A Python platform for distributed formation control of nonholonomic robots. Consensus MPC on feedback-linearised dynamics, solved as a Second-Order Cone Program at 60 Hz.
Simulation and hybrid control of a cart-mounted double inverted pendulum across all four equilibrium configurations, with a hardware build under way.
A battery-powered buggy that follows a track, climbs an 18-degree incline, turns 180 degrees on a Bluetooth command and stops at the end of the line. Three PID loops on an STM32.
An Arduino buggy that solves mazes on its own, using ultrasonic sensing and a PID loop.
An autonomous poker system on an SO-101 robot arm. Dual-camera vision, a ROS 2 control stack and game-theory decision making.
First place. The only interface to the QArm was direct joint-angle control, so we derived the inverse kinematics geometrically and validated it on a digital twin before touching the real arm.
Best team solution. A distributed algorithm that gathers any number of randomly moving robots into a single cluster.
First place in the algorithmic-trading track with a market-making bot. The win took 60 percent of the prize pool.
Second place. A vision pipeline that drives a robot through a line maze from one overhead camera, planning the shortest path and steering in real time.
Third place. A Vue.js app that pulls live football data from an API and presents it in real time.
Coordinating up to three MONA robots from a shared arena camera to push a trolley across the arena. The system worked in simulation, and hardware faults kept it off the arena floor.
A year in industry. Implemented Google's Snappy compression in FPGA hardware with a 10 percent better ratio and 30 percent higher average speed, and redesigned a duplicate MAC-address search for an order-of-magnitude efficiency gain.
Adaptive impedance control for nuclear decommissioning, deployed on 7-DOF KINOVA and Franka arms with formal stability and safety guarantees.
The University of Manchester's award for exceptional all-round performance, placing in the top 0.5 percent of the graduating cohort.
First prize for the best overall performance in the cohort, and the prize for the best final-year project, EEE Department, University of Manchester.
Recognised among the top 10 students across the Faculty of Science & Engineering for academic achievement.
A merit scholarship from the Institution of Engineering & Technology for high-achieving students on IET-accredited degrees.
The University of Manchester's flagship social-responsibility award, earned alongside the degree.
A fully-funded doctoral studentship for PhD research at the University of Manchester.
Cambridge Outstanding Learner Award for the highest mark in Nigeria in Cambridge International AS-Level Physics.
University of Manchester. Safety-critical Control Barrier Function methods for multi-agent autonomous systems. Fully funded.
University of Manchester. First-Class Honours, top 0.5 percent of the cohort. Dissertation on safety-critical consensus-based formation control of distributed mobile robots.