Remote Inspection Optimisation with Heterogeneous Mobility Platforms

Lead researcher

Jingyang You

Project based at

Lead Partner Organisation

The objective of this research is to develop scalable computational methods and software to compute good strategies to inspect large scale assets, despite the presence of information gaps and unfaithfulness of the 3D models of the assets. The inspection strategies will be executed by a mobility platform, which can be a human surveyor or a robot, equipped with one or more visibility sensors. Given the surveyor and the above problem, we refer good inspection strategies to be strategies that are: 

• Robust against the errors, information gaps, and other types of uncertainty of the 3D models, 

• On average, lead to high coverage of the entire assets, 

• Asymptotically optimal in terms of expected path length or scanning cost, 

• Adaptable to potential changes in the environment, 
• Can be computed on a laptop on-board of a mobility platform (e.g., intel i7 with 64GB RAM) within reasonable time. Pre-computation is acceptable, but adaptation of the strategies to potential changes must be computable on-line in at most several minutes.
   

Expected Impact: 

The purpose of this research is to improve the reliability and efficiency in inspection of large-scale industrial assets. Such inspection capabilities will reduce costs of surveys, while improving the data quality for constructing the asset’s digital twin. In turn, such improvements will help improve the safety and longevity of the asset of interest.