Emerge Lab

The Emerge lab at Mercer Labs. On the right, the view of a car driving around with a rendering of its partially observable view. On the left, a visualization of the entire scene. Video of the type of stop and go wave that we've worked on eliminating.

Research

The mission of the EMERGE lab is to make it easy and efficient to develop capable, safe, and intelligent multi-agent systems through learning, simulation, and data. We emphasive use-inspired basic research where we draw inspiration as to what research is needed from specific problems in multi-agent autonomy and transportation. Right now some key interests are:
  • Autonomous validation: how can we create closed loop validation systems that continually find and fix agent failures.
  • Building human-like agent models to improve our simulation and safety capabilities.
  • Scaling up MARL with GPU-based simulators.
In the past we have:
  • Built traffic-smoothing cruise controllers that were deployed at scale in field testing.
  • Showed that simple MARL algorithms scale very nicely.
  • Built fast simulators for autonomous driving and traffic control.

News