Autonomous Vehicles & Artificial Intelligence Laboratory (AVAILab)

Multidisciplinary laboratory focused on developments and applications within the fields of Mathematical Modelling, Optimisation, Soft & Natural Computing, Self-Organisation & Swarm Robotics, Autonomous Navigation and Positioning Systems.

Pilot Research Projects

1)  Light-Tracking Drone Fleet

Lead ResearcherDr Mauro S. Innocente

  • This is a rather academic problem aimed at testing algorithmic developments for fully autonomous swarm-intelligent drone fleets.
  • Drone’s self-organisation is based on a memoryless particle swarm algorithm.
  • Red dot: Drone holding the location of the lightest spot currently found.
Fig. 1: Swarm of forgetful particles following a randomly moving source of light.


2)  Microscopic Traffic Modelling Using Cellular Automata

Lead ResearcherDr Mauro S. Innocente
Contributing Student: Luke J. Smith (MSc student, 2019)

  • Based on Nagel and Schreckenberg’s model (NaSch), with additional features incorporated.
  • 1D model. No junctions, two lanes, one traffic light.
Fig. 1: Variables description.
Fig. 2: Schematics for two-lane model.
  • A two-lane road is simply two single lane roads which have an interaction before the update procedure.
  • The data for each road is passed through a set of lane changing conditions and vehicles swap lanes if they are satisfied.
  • Once all checks are made regarding lane changes, each lane is then updated independently.
  • Parameter bn stands for distance from vehicle to traffic light.
  • Every car will always try to have vn = vmax:


  • There is a random probability (Pslow) that a vehicles’ speed will be decreased by one unit.
  • Each vehicle is moved forward by vn cells.
  • Circular Boundary Conditions.
  • Probability (Pnew) that a new car will appear in first cell.
  • Pnew can then be used to control density of traffic in simulation.
Fig. 3: Two-lane simulation with a traffic light.

%d bloggers like this: