๐Blog: https://t.co/7iJzdE4WLH
Robots are deploying faster than ever, but how we evaluate them hasn't kept up. The challenge: rigorous evaluations are costly, end-to-end foundation models are black boxes that are hard to unit-test, and our benchmarks may not line up with reality.
๐ช So we put together an actionable guide. Whether you're writing a paper, prepping a demo, or working toward deployment, it offers practical frameworks, tips, and resources to:
๐ Plan your evaluation strategy (from blue-sky research to safety-critical deployment)
๐ Build confidence when resources are tight
๐ Connect your evaluations to real-world needs
How should we evaluate robots in the age of foundation models?
We hosted the RSS RoboEval Workshop with folks from academia, industry, and policy to discuss this.
๐ช We put together an actionable guide and insights to get you started on robot evaluations. Link below.
๐งต1/N
A great summary of our work on AI-powered robot navigation aboard the International Space Station: https://t.co/WD1rEJmvlj
@StanfordEng@StanfordAILab@NASA
๐ First demonstration of learning-accelerated trajectory optimization in space! Our team used a neural network to warm-start trajectory optimization for the @NASA's Astrobee free-flying robot on-board the @Space_Station, cutting solver iterations by up to 60% while maintaining safety constraints.
๐ฅ Video: https://t.co/t0OFR8fHBl
๐ Paper: https://t.co/ohBE1iowOB
๐ Submitted to iSpaRo 2025
With @SomritaBanerjee and @ACauligi
We won the best paper award at the AI4Space Workshop! Here's our framework for how ML models can *detect* and *adapt* to changing input distributions, using OOD detection, subsampling, and continual learning. https://t.co/29IEJYduq9 #eccv@drmapavone@StanfordASL@AerospaceCorp