Research is a passion of mine and as a faculty member, I have been building a robotics and automation research program that focus on in situ sensing and control for various unstructured / adverse environment conditions.
Thrust (1) Environment: unknown / adverse in-door conditions
LiDAR and cameras are frequently used as sensors for simultaneous localization and mapping (SLAM). However, these sensors are prone to failure under low visibility (e.g. smoke) or places with reflective surfaces (e.g. mirrors). On the other hand, there are other ranges in the electromagnetic spectrum that exhibit penetration properties due to larger wave lengths and thus not affected by low visibility. Hence, my group has explored novel sensors like UWB Radar, mmWave and WiFi to enable the robots to operate in adverse conditions as illustrated in the following Figure which is applicable for users like firemen.
To enable localization of officers as well as robust odometry of robots, we have worked on AI-Aided IMU to out odometry. In addition, various exploration algorithms are designed to enable the robots to perform search and rescue in unknown building. As seen in the figure, we have also used other novel sensors like event-based for high-speed moving robots.
Thrust (2) Environment: Healthcare environment without installing infrastructure
Another environment that I have special interest is the healthcare settings. Automation for healthcare institutions is interesting because it involves both patients as well as the staff. The environment is often crowded and often under manpowered. Hence, my group has taken up several healthcare automation projects and happy to make many friends. Two of the selected projects are shown in the following figure.
Thrust (3) Environment: Vibration or experiencing disturbance
Vibration (eg: hand tremor) / disturbance (eg: strong wind) and precision usually do not co-exist. However, there are several cases whereby precision is still required as shown in the following figure. One example is surgeries. There are several distance like tremor and involuntary respiration motion. Yet, precision is required during the surgery. Hence, my group has proposed various solutions in this area. Another example of source vibration / disturbance is when trying to perform targeting of a distant object on a moving vehicle. The engine and road introduce disturbance which make precise distant targeting challenging. My group has managed to use the algorithms delivered in the surgical tools to enhance and applied onto this area. Similarly, my group has also been funded by ST Aerospace directly to work on enabling a vision system on mast to perform defect detection out on the airport tarmac (ie not in a hanger).
Thrust (4) Environment: Humans
Sensing the physiological status (eg: Cognitive load, situation awareness) is an interesting area as the natural behaviour of the human cannot be structured. In this area, we have been using time series foundation model to estimate the status as shown in the following figure.
External Funding Sources
| Title | Role |
| Event based Camera for Image Stabilisation | PI |
| Predictive Control for Optically Stabilization of High Frequency Vibrations | PI |
| Upscaling of 3D Food Printing and Culture of Fish Fillet | PI |
| Project NEMEAEUS | Task PI |
| Project Procyon | Task PI |
| 3D Localisation for UAV with 5G Radio Fingerprinting | PI |
| Robotics Middleware Framework 2.0: Enabling large-scale robotics systems in multiple sectors – Healthcare | Co-PI |
| Nurse-Robot Teaming- Autonomous Patients’ Excreta Disposal and Monitoring | PI |
| Transforming waste recycling with multi-spectral AI technologies | Co-PI |
| Development of Robotic Indoor Tracking System | PI |
| Depth-Vision based Localisation for Urban Air Mobility | Co-PI |
| BiPap for A&E Patients | Technical PI |
| Beta Draconis: Manned-Unmanned Teaming with Multiple UGVS | Task PI |
| Sensor Alert System for Surgical Drainage Bottles | Technical PI |
| Mobile collaborative robot platforms with vision feedback and adaptive and learning soft end-effectors for coning/de-coning wharf operations | Co-PI |
| Project Coriolis: Multi-Modal Interaction Research and Simulator Development | PI |
| Next-Generation Brain-Computer-Brain Platform – A Holistic Solution for the Restoration & Enhancement of Brain Functions (NOURISH) | Co-PI |
| Precision localization of payload for external aircraft inspection | PI |
| Development and POC of Nurse-Robot Teaming for Care Delivery | PI (SUTD Portion) |
| Enabling digital gastronomy for 3D food printing | PI |
| Intelligent Ground-Aerial Robots for Closed Quarters Interactive Operations | Co-PI |
| Industrial Digital Design and Additive Manufacturing Workflows | Co-PI |
| Project Borealis: Indoor Manned-Unmanned Teaming with Multiple UAVs | Task PI |
| The smart handrub bottle | Co-PI |
| Tweaking everyday objects to prevent the spread of AMR | Co-PI |
| Evaluating the effect of a bedside rehabilitation chair on functional decline in hospitalized older adults | Co-PI |
| Localization with Collaborative LTE and Wifi SLAM | Co-PI |
| Automated Part Extractor for a Fleet of Fused Filament Fabrication (FFF) Printers | PI |
| Automated Part Extractor for a Fleet of Fused Filament Fabrication (FFF) Printers | PI |
| Human-Robot Collaborative AI for Advanced Manufacturing and Engineering | Co-PI |
| Human-machine interface and technology development for collaborative control of unmanned systems | PI |
| Kyanite – Improved Cognitive Performance by Design | Task PI |
| Privacy Panels to Improve Hand Hygiene Rates | Technical PI |
| Integrated Physics &Math E-Learning | PI |
| Telamon – Indoor Relative Positioning System (IRPS) | Co-PI |
| Feasibility study of the sensing capability for vibration control | PI |