Embedded Sensing, Communications and Processing Lab
Keywords: Learning based Robots, Collaborative Robotics, Energy Efficient Hardware Accelerators, Autonomous Intelligent Machines <We are hiring>
How do we make mobile Internet of Things that are autonomous, safe & secure, taking advantage of the latest advances in Smart City Communication technologies?
An open API middleware to support data exchange between data producers and consumers, that is secure, privacy preserving, supports video and distributed secure analytics. Enable new smart city applications like correlating traffic and pollution, coordinated emergency response across fire, water board, hospitals and traffic police, etc.
What is the general framework to make any multi-legged robot to navigate different terrains? We are exploring data driven approaches like Deep Reinforcement Learning, to make our pet "Stoch" traipse across our campus and chase away stray dogs. See a video.
We suspect it might be easier to teach robots than students! So we are working with Yaskawa, Japan to explore concepts from meta-learning and imitation learning to teach robots to work in unstructured environments.
High resolution air quality monitoring and air pollutant data analytics
Can we make reliable low cost pollution sensors? How do we overcome their deficiencies algorithmically? Can we do pollution sensing using traffic cameras?.
Vibration powered Sensor Interface chip
Can we make a completely wireless sensor ASIC for our rockets and satellites? We are exploring multi-modal energy harvesting in combination with back-scattering based communication technologies to achieve this.