• Lighting R&D Intern in Daylighting Lab, Osram Sylvania

  • The term “daylighting” is used in various ways, on one hand for harvesting daylight to illuminate spaces, and on the other, more recently, for using artificial light sources to replicate daylight patterns. The emergence of solid state lighting (SSL) opens up a large number of new avenues for daylighting. The combination of natural and artificial lighting could improve energy savings, and increase comfort of living beings. The Daylighting Lab is set up in the Osram Sylvania research facility building. This PhD intern project focuses on the development of platforms with algorithms and architectures for a building-wide lighting control system involving sensors, actuators, LED lights and daylight regulating devices. This entails embedded system design to read raw data from sensors, process and analyze the input data, and control DALI-networked lighting equipment. Advanced features are designed to improve the quality of light by adaptive automation of intensity, color temperature, and color mixing in closed-loop manner. In the future buildings, comprehensive systems must connect networks of energy distribution, lighting, heating, cooling and ventilation.

  • Wireless Video Camera Networks for Coastal Erosion Monitoring

  • Still images and video are important media to observe rare and extreme events, and to document both short- and long-term changes in ecology and environment. This project specifically investigates a wireless network of web cameras to monitor shoreline erosion and wave dynamics. Such network is designed to operate long-term, low-cost, low-power with energy-harvesting in the field, and provide remote access and live video. All devices are weatherproof packaged, powered by solar energy, and IP-enabled. Smart controllers of Arduino and Real-Time-Clock shield are designed and deployed to synchronize all devices on/off to high tides everyday. This wireless network of cameras has been functioning as designed since installation on Thompson Island, Boston Harbor in September 2012, and is expected to operate with low maintenance. Coastal erosion has been captured in exciting images and videos during Hurricane Sandy and winter snow storms. This network is capable to produce in quantity and implement at scale as an ecosystem observation platform.

  • Networking the Physical World

  • Microsystems like Arduino, Sun SPOT and Tiny OS platforms enable implementation of wireless sensor networks. This project explores IEEE 802.15.4 (physical layer of Zigbee) mesh networks for remote access, localization, routing, distributed discover and leader election in swarms. Onboard sensors (light, temperature, accelerometer) and GPIO interfaced sensors and actuators (PIR, ultrasonic range finder, IR proximity, compass, microphone, speaker, servo, ESC, etc) are utilized to remote control a 1/8th scale electric crawler, and drive autonomously.