From Hacking to Programming: Lessons Learned in Creating Assistive Environments for the Elderly Sumi Helal, Jeffery King, Hicham Zabadni, Raja Bose, and Hen-I Yang The Pervasive and Mobile Computing Laboratory Computer & Information Science & Engineering Department University of Florida, Gainesville, Florida 32611, USA http://www.icta.ufl.edu Abstract The first generation of pervasive systems focused on demonstrating their feasibility and benefits in various application domains. They mainly relied on ad-hoc system integration—interconnecting sensors, actuators, computers, and other devices in the environment. Most of these prototypes were for experimentation only; they are certainly unsuited for commercial deployments. They were difficult to deploy, had issues with scalability, lacked an ability to evolve, and had a high cost associated with the insertion of each of their constituent components [1]. In this paper we present our experience in building a first generation “assistive environment” – a pervasive space designed for older and disabled people. We zoom in on several important lessons we learned and show how they influenced our view (and shaped our research agenda) of what a second generation pervasive system should or could be. Keywords Pervasive Computing, Smart Spaces, Programmable Pervasive Spaces, Distributed Systems, System Integration, Sensor Networks, Assistive Environments 1. Introduction The Pervasive and Mobile Computing Laboratory at the University of Florida has been researching and prototyping “assistive environments” – pervasive spaces designed to increase independence and improve quality of life for older people and people with special needs. The research and development activities have been done in close collaboration with researchers from UF’s School of Public Health and Health Professions. The main objective was to create a usable and acceptable technology that: (1) supports aging in place, and (2) augments age- related cognitive and motor impairments. While the objective was crystal-clear, achieving it was very challenging but educating at the same time. Matilda Smart House (Figure 1) was our first attempt at creating assistive environments for the elderly. Housed inside the pervasive computing lab and occupying over 500 sq ft, the mockup house consisted of a front door, a kitchen area, a living room area, a small bedroom and a tiny bathroom. It was a platform to experiment with accessible appliance control, indoor location tracking and cognitive assistant concepts and applications. We integrated tens of sensors, actuators, appliances and other components including contact sensors, motion sensors, cameras, ultrasonic transceivers, X10 modules and controllers, several microprocessor-based controllers, a microwave oven, an entertainment system, mobile phones, and a home PC. The complexity of the integration task in Matilda Smart House far exceeded any expectations we ever