Emerging robotics applications have heavy computational burdens, and must perform tasks at real-time rates while under strict power budgets. Traditional CPU-based solutions are unable to provide the performance and energy efficiency to satisfy these demands. To address this challenge, we are (1) developing software libraries that make it easy for robotics researchers and practitioners to use alternative computing platforms (e.g., GPUs and FPGAs), as well as (2) developing automated tools to enable the efficient design and use of custom robotics accelerator chips.
