Embedded Research Lab
Embedded systems are used everywhere – starting monitoring meters to mobile devices and complex robots or vehicles. The complexity of such systems is increasing at an exponential rate due to both advances in technology and demand for realization of ever more complex applications in the areas of communication, control, multimedia, entertainment and even human intelligence.
ITU Embedded Research Lab is focusing on developing efficient techniques,methodologies, frameworks and prototypes for low-power, high-performance and reliable embedded systems and Internet of Things. Currently, we are investigating many exciting and cutting-edge research problems including hardware/system security and trust, Distributed Energy Management System, Smart Li-ion Battery Pack Design, Voice control for IoT/Robotics and hardware/algorithm co-design based on FPGA/ASIC.
Distributed Energy Management System
Dr Karl Wang, Kevin Huang, Peter Guan, Adam Zeng, Jinhui Li, Jingya Xu
Green home, building or campus design and construction can generate important energy savings and greenhouse gas reductions. To achieve these gains, it is necessary to optimize the way energy is managed. A realtime and two-way electricity meter is the key for energy monitoring, prediction and control. And In the future, energy management will become more complex due to the integration of renewable energy production, energy storage as well as the ability of buildings and machine to respond to signals from the electrical grid during demand response events. The high performance buildings and factories of tomorrow will have to adjust their operating conditions in real time in response to their equipment conditions, predicted energy consumption and peak load management scenarios.
To establish the basis of innovative control systems for future campus and factories, We are conducting research and development on energy management and control strategies and prototypes in buildings. These research activities focus on the following subjects:
- Smarter meter and low power sensor design and integration
- Energy monitoring and prediction
- Development and implementation of smart analysis and control strategies for demand response in commercial buildings and
- 4 patents
- New Sensor design
- Green campus project
- Distributed IoT platform and APIs
- Finish and demo the Green Campus Project
- Data mining and analysis
- Exploring the industrial application for machine monitoring and control
Smart Li-ion Battery Pack Design
The battery system is still the important and expensive element in today’s mobile devices, electric vehicles and energy storage. The battery cell charging, monitoring and management concept and prototype developed by Embedded System Lab allow a significant reduction of the final battery pack costs, enable the scalability and manageability both at local battery pack design and at cloud.
- The monitoring and charging functions are integrated on each smart battery cell/pack, thus improving the reliability, the assembling and contacting efforts.
- Good modularity and flexibility of the usage of the smart battery cell/pack.
- The hardware of the battery management system can be standardized and reused in new and different developments by improving only the software part.
- The smart battery cells/packs provide all the necessary integrated electronics to be monitored.
Voice Control Technology for IoT and Robot
Voice and speech is the primary means of communication among people.The Speech Recognition,Text to Speech (TTS) and Speech to Text (STT) can help people to interact with machine. And we believe that speech related technology and product will become more and more popular in human-to-machine interaction. We are working on leveraging the current speech recognition technology and machine learning to build up the voice control and conversation project for smart IoT and robot projects.
With the technologies and applications advanced, our alarm clocks, thermostats, appliances, and any other device that can connect to the internet is a part of the Internet of Things. These devices use their ability to connect to the internet to communicate information to other devices and people. This makes IoT devices much smarter than regular devices. And it will be much more wonderful if the device can listen and talk to us, communicate and share information with we users in natural, humanlike way.
The adoption for speech technology for robot seems more nature and promising. With voice technologies, not only the robot can be manipulated using simple voice command, but also the sophisticated and lifelike conversational interfaces and interactions can be built to implement the applications using voice and natural language, like conversational bots, personal assistants.
Poster: Yongcan Huang, Min Zeng, Jie Chen, Jinhui Li, Karl Wang, A Low-power Current & Voltage Sensor Using an Optically Coupled Isolator, Sensors Expo & Conference 2017