Challenge
beendet!

Shell Energy and Chemicals Park Rhineland

Fluid Leakage detection via drone systems and Artificial Intelligence (AI)

Detailed description of challenge

  • There are several kilometers long piping in our park which need to be monitored very closely
  • At the moment this is done by our own staff or contractors, which is very time consuming because they need to walk each kilometer
  • Through the limitations of doing this manually, e.g. the pipe tranches are very wide and can not be checked equally, the monitoring quality is poor

First initiatives within company

  • Started a drone program to be able to fly with pre-configured flights above the pipe racks and pipe tranches
  • Regulative stuff is done, so we can fly
  • We did a pilot program with a startup, which mapped the area and tried to implement an AI algorithm to detect corrosion and leakages. AI capabilities of this company are in the early stages and we decided to look for a specialized company

Need for support

  • We need a solution to visually detect leakages (dropping fluid) under pipe racks and pipe tranches
  • We need a solution, which can monitor several kilometers of pipes; in best case fully automated with minimized human interaction
  • We need a solution, which can do this mostly autonomous with a drone (weight limitation <2 kg)
  • The challenge should focus on the development on an AI algorithm, which can “translate” the visual inspection data into an evaluation pattern, which identifies potential fluid leakages of the pipe grid
  • Optional: if there is an alternative to the purely visual inspection (e.g. specific sensor technologies) Shell would be open for this as well