Case study · 01 / 12
Swarm Routing
AI that teaches delivery drones to share the work
- Year
- 2025
- Focus
- Python
- Status
- Complete
Live
My MSc dissertation at Birmingham City University: an AI system that routes swarms of delivery drones, tested from 2 to 50 drones.
One drone following a fixed route is easy. Fifty drones sharing a city, with blocked drop-offs, shifting weather and draining batteries, is not. I built the whole system in Python. I generated delivery points, split the city into zones with spectral clustering, planned a route inside each zone, then trained the drones with Q-learning to adapt when things went wrong. The work won the PGXPO 2025 Research Innovation Award and the Faculty Award for Dissertation Excellence.
What it does
- 01Generates realistic delivery points across an urban layout
- 02Turns the city into a graph of points and distances
- 03Splits the graph into delivery zones with spectral clustering
- 04Plans a route inside each zone
- 05Assigns a zone to each drone in the swarm
- 06Re-routes when a delivery point is blocked
- 07Adapts to weather shifts and battery depletion
- 08Measures completion rate and delivery time for every run
How it was solved
- 01Started with a nearest-neighbour route as the baseline to beat
- 02Partitioned the delivery graph with spectral clustering so drones stop competing for the same points
- 03Trained drones with Q-learning over 500 episodes, then scaled to multi-agent reinforcement learning
- 04Added disruptions mid-run to test how the swarm recovers
- 05Scaled from 2 to 50 drones to prove it holds up
30% higher task completion than static routing, 25% shorter delivery times, and a further 15% gain in completion under disruption.
Stack
- Language
- Python 3.10
- Libraries
- NumPypandasSciPy (spatial distance)scikit-learnMatplotlib
- Routing
- Travelling salesman problemNearest-neighbour heuristicDistance matricesGraph adjacency matrixPer-zone re-solving
- Clustering
- Spectral clusteringPrecomputed affinity graphk-means label assignment
- Reinforcement learning
- Tabular Q-learningEpsilon-greedy explorationCustom training environmentMulti-agent RL (independent learners)Reward shaping
- Simulation
- Seeded delivery-point generationBattery and recharge modelNo-fly zonesBlocked-point disruptions
- Tools
- Google ColabJupyter Notebook
- #ai
- #reinforcement-learning
- #drones
- #python
- #research
That is the work. Now meet the person who built it.