As many robot automation applications increasingly rely on multi-core processing or deep- learning models, cloud computing is becoming an attractive and economically viable resource for systems that do not contain high computing power onboard. Despite its immense computing capacity, it is often underused by the robotics and automation community due to lack of expertise in cloud computing and cloud-based infrastructure. Fog Robotics balances computing and data between cloud edge devices. We propose a software framework, FogROS, as an extension of the Robot Operating System (ROS), the de-facto standard for creating robot automation applications and components. It allows researchers to deploy components of their software to the cloud with minimal effort, and correspondingly gain access to additional computing resources and predeployed software made available by other researchers.

To accommodate the real-time update requirements for many Machine Learning models and robotics applications, Feature Stores are fast emerging as a new class of Machine Learning system that maintains intermediate statistics of live data streams used for model training and inference to improve accuracy and save prediction time. Our work is based on RALF, a feature store designed for streaming data and explicitly leverages downstream feedback. Our project explores the impact of lazy evaluation, which postpones feature updates in a feature store, on the three most important aspects of feature stores (i.e., staleness, latency, and costs) and builds an SLO-aware featurization scheduler that reduces the staleness of the queried features by co-scheduling feature updates and query responses.




Download Full History