Full Workshop 7

Mars 3.0 – Martian Autonomous Robotic Swarm

MARS 3.0 will develop a goal oriented autonomous robotic system based on multiple agents. The final aim is to design or generate behavioural rules for a group of robots that would eventually lead to collective fabrication of a Martian habitat structure. This behaviour will be developed and tested both in virtual and physical simulation. Swarm intelligence is one of the main focuses of the workshop. Participants will explore the definition of rules for interaction between robotic units as well as their surrounding environment to accomplish the goal. It is a multidisciplinary effort where mechanics, electronics and software disciplines will come together to accomplish an architectural challenge in the field of behavioural machines. The system will operate in a 2.5D unknown and changing physical environment without human intervention. The challenge has redundancy at heart, i. e. autonomous units that can adapt to change and failure without resulting in mission failure if one of the units malfunctions.

Sponsors:

Josef Musil (Electronics and Algorithms, Foster + Partners), Miriam Dall’Igna (Electronics and Algorithms, Foster + Partners – UoW, OU), Xiaoming Yang (Electronics and Algorithms, Foster + Partners), Reinier Zeldenrust (Electronics and Algorithms, Foster + Partners), Octavian Gheorghiu (Media and Algorithms, Foster + Partners – UCL, AA), Samuel Wilkinson (Additive Manufacturing and Algorithms, Foster + Partners)

The participants will have a chance to engage in behavioural programming by developing their own algorithms or help evolving the existing ones. The aim is to set up the hardware in a manner that diverse algorithms can be tested with the same electronics/communication setting out. In addition, there will also be other technical challenges related with sustainability such as energy expenditure, networking to enable communication between robots as well as autonomous landscape mapping and navigation using SLAM.

Previous experience programming required. Python is preferred and will be used during the workshop, but knowledge of other languages will be beneficial.

Experience with Linux and ROS encouraged. Knowledge of the swarm behaviour programming, networking, communication and electronics will be beneficial.

3 days

10