Real-time Multi-modal Sensor Integration for Robots in Unstructured Environments

The development of safe and efficient robotic sensing and perception algorithms are pivotal to the evolution of a wide range of areas involving robot intelligence, monitoring, and human-robot interaction. This development will improve the quality-of-life and work conditions across multiple industries and settings, including in our own homes and in
nature. One of the biggest challenges of this topic is designing a system that will respond instantaneously and even proactively to various circumstances. Developing robots that work and collaborate safely in unstructured environments is extremely challenging and raises an important question for the design of high-frequency robotic systems: how can we obtain realtime solutions for systems in dynamic environments? We must look at combining the capabilities of multiple dimensions such as vision, haptic, and environmental sensing to enable a robot to navigate successfully in unknown and potentially harsh environments - and this is an area which I would like to investigate. The fields of robotics and sensors have enjoyed tremendous progress over the last decade, both in theory and practice. However, many challenges remain in safe and efficient sensing, perception, and collaboration. How can we ensure that robots can work and interact safely with humans, with other robots, and with their surroundings without compromising efficiency? How will the robot's perception influence its action? How should robots manage unexpected contacts and impacts? These questions are important in safeguarding the continued progress of robotics and sensors, and pose exciting new research challenges for the field.