An international consortium associating Canadian and German partners aims to develop a virtual learning methodology to train robots for operations and maintenance to reduce costs, improve quality, safety, and faster adoption of new advanced materials in aerospace.

A slow production rate, the rapid growth of air transportation and enormous backlog of new aircraft orders make the aerospace industry linger on traditional practices and prevents it from moving fast enough to adopt more efficient aircraft designs and advanced materials.  

An increased level of automation via the use of robots in manufacturing new aircraft and maintenance, repair, and overhaul (MRO) of the existing fleet is considered a possible solution for cost reduction and improved quality and safety in the aerospace industry. However, traditional industrial robots used in assembly lines of the automotive industry and electronic devices are inadequate for the aerospace industry because of small batch sizes, large components, diversity of products and high complexity and variation in operations. Thus, the current practice of programming or teaching a robot for every specific task is limited, if not futile, in the aerospace industry.  

In Industry 4.0, robots are intelligent, highly adaptive, and trained through machine learning to handle different equipment, tools, products, and materials without a need for explicit programming. However, machine learning requires a large volume of data for capturing all possible physical experiences to train the robot, which can be too expensive or unavailable. Recent advances in robotics demonstrate the feasibility of learning from synthetic robot experiences and simulations.  

In the AIARA project, we aim to develop a methodology to use learning results from simulation and virtual environments by means of so-called Reinforcement Learning to train real robots for a wide range of aerospace manufacturing processes and MRO operations. We will evaluate and demonstrate the feasibility of this approach using four benchmarking use cases including:  

  • a draping robot for composites manufacturing; 
  • a multi-arm robot for handling of flexible material in composites manufacturing; 
  • a robot is trained in a virtual environment so that it can handle and refill a 3D printer in way that usually requires the manual work of a human; 
  • a test bench that shall be implemented where a robot measures the electrostatic forces of an end effector where the influence of various environmental parameters shall be analyzed by means of machine learning. 

This research partnership brings together the UBC, Element AI and Kinova from Canada side and German Aerospace Center (DLR), Broetje, Fraunhofer IPT and ZAL (Center of Applied Aeronautical Research) from Germany to offer a more productive path for the aerospace industry under Industry 4.0. 

The funding bodies are: 

  • Bundesministerium für Wirtschaft und Energie (BMWi) in Germany 
  • Ministère de l’économie et de l’Innovation du Québec (MEI) in Québec  
  • Natural Sciences and Engineering Research Council of Canada (NSERC) in Canada 

The international collaboration project AIARA is in line with the Montreal-Hamburg partnership, which brings together Hamburg Aviation, ZAL Center of Applied Aeronautical Research and the Consortium de recherche et d’innovation en aérospatiale au Québec (CRIAQ), Aero Montréal with the aim of building successful collaborative relationships between German and Canadian expertise in collaborative research.