Ion Stoica
Papers
8
Total Citations
225
H-Index
6
About
Ion Stoica is a pioneering researcher in cloud and fog robotics, with a focus on enabling robots to offload computationally intensive tasks to cloud and edge computing resources. His major contributions include the development of FogROS2, an adaptive platform that leverages cloud computing (AWS, GCP, Azure) to run contemporary robot algorithms at desired rates despite mobility and power constraints. He has also advanced distributed motion planning using serverless lambda computing, as demonstrated in his work on Fog Robotics algorithms (21 citations). In reinforcement learning, Stoica introduced Deep Discovery of Options (DDO) and Multi-Level Discovery of Deep Options (70 citations), which augment agent control with higher-level behaviors to reduce sample complexity in high-dimensional state spaces. His hierarchical imitation learning for home automation (24 citations) reduces the number of demonstrations needed for robot learning. More recently, he has explored composing MPC with LQR and neural networks for amortized efficiency (15 citations), and fault-tolerant cloud robotics through FogROS2-FT (2 citations). With over 225 total citations across his most-cited works, Stoica's research bridges the gap between robotic autonomy and scalable cloud infrastructure, making him a leading figure in the field of cloud robotics.
Research Focus
Key Achievements
Top Papers
- 1Multi-Level Discovery of Deep Options70 citations · 2017
- 2
- 3FogROS2: An Adaptive Platform for Cloud and Fog Robotics Using ROS 238 citations · 2023
- 4Multi-Task Hierarchical Imitation Learning for Home Automation24 citations · 2019
- 5
- 6
- 7
- 8FogROS2-FT: Fault Tolerant Cloud Robotics2 citations · 2024