Projects

Hardware Acceleration for Robotics

The next generation of robotics applications are facing a computational crisis. Facing thermal dissipation limits, CPUs have hit a single-threaded performance wall. At the same time the real-time computational demands for emerging robotics algorithms are only growing. To alleviate this bottleneck, we are developing software libraries that make it easy for other robotics researchers and practitioners to use alternative computing platforms (e.g., GPUs and FPGAs), as well as developing automated tools to enable the efficient design and use of custom robotics accelerator chips.

The Tiny Machine Learning Open Education Initiative (TinyMLx)

TinyML is a cutting-edge field that brings the transformative power of machine learning (ML) to the performance- and power-constrained domain of embedded systems. Successful deployment in this field requires intimate knowledge of applications, algorithms, hardware, and software. We are working to improve global access to TinyML educational materials through our MOOC courses, the TinyML4D Academic Network, and a variety of other projects.

Robust Nonlinear Trajectory Optimization

It is often challenging to get nonlinear trajectory optimization to converge reliably without getting stuck in spurious local minima. Better methods for handling constraints in the optimization problem and improved design of objective (cost) functions can help alleviate this problem. Leveraging new methods of data-efficient machine learning and mathematical insights from optimization theory, we are working to design robust trajectory optimization methods that can be used for real-time nonlinear model predictive control.

Tiny Robots

Many emerging robotics use cases will require small, cheap robots that use embedded devices for computation. When compressing robotics algorithms to fit on these resource constrained computational devices, new challenges and opportunities emerge. We are working to unlock the full potential of these tiny robots by leverging insights from both computer architecture / embedded systems and robotics to custom tailor algorithmic solutions through hardware-software co-design.