During my Internship at Ford’s Innovation Lab in Silicon Valley, I developed a framework for scalable multi-agent simulation.

This C++ based framework allowed to simulate hundreds of vehicles at the same time in order to genereate realistic traffic scenarios. While the majority of the vehicles are actors, it is possible to run multiple egovehicles with custom path-planning, decision-making, and controls algorithms. These vehicles could then be virtually driven for thousands of miles to identify edge-cases where the algorithms failed.

During this internship, I was able to practice and improve my C++ skills, used differential equation solvers, and got very confortable working with Protocol Buffers (Protobufs).
Furthermore, I got introduced to parallel programming on Ford’s High Performance Cluster.
I also got more confident dealing with the challenges of synchronization, parallelization, and vehicle dynamics.