MarketWatch, May 4, 2021. Caspar.AI uses the whole home as a sensor – already existing elements like sensors on switches and appliances. From automated control of lights, shades and temperature to more complex tasks such as fall detection, the Caspar system works without cameras and without “wearables” such as resident pendants.
Learning to Represent Haptic Feedback for Partially-Observable Tasks. ICRA, 2017.
Physical environments such as automated homes need to be able to take actions even in presence of limited, incomplete observations.
We formulate the task as a POMDP and learn an appropriate representation with deep recurrent neural networks that serves as the state for the POMDP model.