The goal of the Physical Analytics project, or Phytics, is to perform analytics on the physical actions of users. Such analytics would be valuable in a variety of contexts where the design of a physical space is intertwined with how users use the space, e.g., to understand the behaviour of shoppers in mall or a store, which might be indicative of their interests; to fine-tune the layout of services in an airport or a train station based on how people pass through; and to plan the layout of exhibits at a tradeshow based on the physical browsing patterns of users from previous editions. Thus, Physical Analytics brings to the physical world capabilities akin to those taken for granted in the online world, where the ready availability of clickstreams drives rich analytics.
Our research in realizing Physical Analytics spans several problems, including multi-modal sensing to track user activity; distributed data processing, spanning elements resident in smartphones, the local infrastructure, and the cloud; and interfacing with users through incentives for participation and targeted information delivery.
- Swati Rallapalli, Aishwarya Ganesan, Krishna Chintalapudi, Venkata Padmanabhan, and Lili Qiu, Enabling Physical Analytics in Retail Stores Using Smart Glasses, in ACM Mobicom 2014, ACM – Association for Computing Machinery, 7 September 2014.
- Rajalakshmi Nandakumar, Swati Rallapalli, Krishna Chintalapudi, Venkata N. Padmanabhan, Lili Qiu, Aishwarya Ganesan, Saikat Guha, Deepanker Aggarwal, and Aakash Goenka, Physical Analytics: A New Frontier for (Indoor) Location Research, no. MSR-TR-2013-107, October 2013.
Workshop on Physical Analytics at ACM MobiSys 2014 (June 2014)