The SPIRIT Center performs systems research with the objective to design & develop a large-scale, complex smart systems platform for translational R&D of technologies relevant to smart nation applications. Some of our core research areas are highlighted here.
Smart Nation Technology Systems Research
Smart Nation/Cities is enabled by the widespread adoption of new technologies such as Sensors, Internet of Things (IoT), Cloud computing, Mobile technology, Big data analytics and Cybersecurity to develop intelligent systems in order to improve government operations, support better living, create opportunities, and to support stronger and safer communities.
Judiciary Collaboration: AI to aid Courtroom Procedures
To improve judicial efficiency through AI techniques we pursue R&D for an Intelligent Case Retrieval System with Decision support tools for the judges and legal professionals, for developing advanced machine learning algorithms, intelligent information retrieval and natural language processing techniques; and supporting efficient retrieval of precedent cases, perform automatic summarization and legal reasoning.
MITRE Smart Nation Collaboration: Worker Health and Safety
To improve construction site safety through predictive analytics, we fuse diverse data sources from low-cost smart sensors to detect “near-miss” events in construction sites. We develop advanced machine learning algorithms to identify patterns and generate predictive analytics. And we generate data visualization dashboards which allow decision makers to leverage these new data sources to improve safety planning and mitigate risks.
Video Compression and Secure Transmission
To enable video compression and secure video transmission for elderly care, we designed secure and lightweight compressive sensing technology using stream cipher, published in the IEEE Transactions on Circuits and Systems II. It achieves 2x – 40x compression with tunable quality. It has a lightweight design and it suitable for sensor nodes. Software/Hardware prototype is available and it can be smoothly plugged into existing systems.