Towards Industry 4.0/5.0, a strong technologic shift demands autonomous systems capable of intelligent perception, navigation and action in complex, dynamic manufacturing environments.

Advanced manufacturing is fast evolving towards the "Factories of the Future," characterised by flexible, dynamic reconfigurable work cells, human-robot collaboration, etc.

This particularly requires autonomous mobile robots (AMRs) and robotic manipulators with an unprecedented level of spatial awareness and adaptability. Therefore, advanced sensing and navigation techniques such as Simultaneous Localization and Mapping (SLAM) in advanced manufacturing process have always been of paramount importance. 

SLAM is a critical technology for mobile autonomy. In advanced manufacturing, it enables autonomous robots to conduct many essential tasks such as logistics and inspections etc. However, in the complex, dynamic, and often feature-less or repetitive environments of modern factories (e.g., warehouses with long aisles, workshops with metallic surfaces, highly dynamic human-robot interactions, or cluttered assembly units, etc.), current state-of-the-art SLAM systems have critical limitations. 

These limitations and challenges have presented fundamental knowledge gaps on SLAM techniques for advanced manufacturing. Robust Multi-Modal Sensor Fusion and AI techniques provide a promising solution to solving these challenges.

Based on Dr Erfu Yang’s previous research work, the ambition of this project is to fundamentally investigate novel resilient multi-modal sensor-fusion SLAM framework specifically designed for advanced manufacturing to fill the knowledge gaps by directly addressing the critical challenge of enabling robust navigation and environmental understanding for autonomous mobile robots (AMRs) and robotic manipulators in these settings and research need identified above.

The overall research aim of this project is to design, implement, and validate a novel intelligent multi-modal sensor-fusion SLAM framework that delivers robust, accurate, and context-aware spatial perception for autonomous systems in dynamic, large-scale manufacturing environments, especially   complex human-robot collaborative manufacturing settings where safety, efficiency, quality and human-robot interactions must be considered. 

It is also critical to understand how to ensure the efficiency and flexibility of the manufacturing process can be enhanced when implementing the developed SLAM framework in advanced manufacturing systems. It is important that SLAM-powered autonomous system is able to work collaboratively in a safe manner around a designated area of production lines, such as automated packaging stations. In a busy manufacturing environment, there can be no errors leading to production down.

By being closely connected and complementary alongside the individual areas of expertise contributed by the project team in Strathclyde and industry partner Kudan Limited UK, the key novelty is the development of robust AI/machine learning algorithms, fast-response multimodal sensing  and novel integration strategy that implicitly consider human knowledge, process uncertainty, dynamic interaction and variation through intelligent fusion of multiple sensors and sources of information when implementing SLAM in advanced manufacturing settings. 

Further information

Interviews with qualified and promising candidates will be conducted on a rolling basis until the position is filled.