In the future factory, operators and robots are colleagues

Sudha Ramasamy at University West leads research that could revolutionize the manufacturing of customized products.

Today, the possibilities for configuring a vehicle according to individual preferences are virtually endless. As products become increasingly customised, it also becomes more difficult to create flexible and efficient production flows.

“At Volvo’s truck production line, almost every vehicle has a unique configuration of equipment. This creates challenges for the operators responsible for kitting – that is, those who pick and assemble the components for each vehicle,” explains Sudha Ramasamy, Associate Professor of Automation at University West.

“To create efficient flows where operators avoid heavy and repetitive tasks, we are developing smart AI-driven robots with advanced grippers. The idea is that they will be able to handle a wide range of components using the same gripper and also communicate with operators in English.”

Safe, efficient and seamless collaboration

The aim of the research is to transform manufacturing processes using advanced AI-based solutions. This will enable safe, efficient and seamless collaboration between operators and robots within a flexible automation system. A key focus is on teaching the robot to identify components and safely grasp, lift and move them.

“We use advanced vision-based methods to train the AI models. In practice, this means that a large number of images of components help the robot recognise them and determine the optimal way to grasp each item.”

Another part of the project involves training the robot to understand and communicate with operators in English. A language model is used to enable natural interaction.

Using large language model

“To ensure smooth collaboration between robot and operator, we are developing LLM-based interaction (Large Language Model) that allows them to communicate in English. This could involve planning tasks or solving problems that arise.”

“One of our biggest challenges is creating a reliable AI solution. AI is evolving rapidly, and we need to continuously adapt our system accordingly. It will also take time before the AI robot can work as quickly as a human operator.”

The project will run until August 2028. By then, a demonstrator will be available at Production Technology Center in Trollhättan, and at Volvo’s research campus, CampX in Gothenburg, showcasing the results.

“Industrial jobs can become more attractive”

“This is an exciting project where we take a broader approach to industrial automation to enable smart manufacturing. I hope it can help change perceptions of industrial jobs. With more sustainable and flexible ways of working, these jobs can become more attractive.”

The research is carried out in close collaboration with Volvo Lastvagnar, Chalmers Industriteknik, and the tech companies Hitachi Energy Sweden, Binar Solutions and Repli5.

“Each partner contributes important expertise in different areas. The research team also includes a PhD student focusing on advanced AI-based grasping strategies for robots. In addition, several students are contributing through their master’s theses.”

The research team will present ongoing developments throughout the project, including at Automation Day on 26 November in Trollhättan.

Contact

Sudha Ramasamy, Associate Professor of Automation, University West

Facts

• The AIKIDO research project is funded by Vinnova and runs until 2028.
• Five partners contribute expertise to the project:
  - Volvo Lastvagnar contributes industrial use cases, providing technical inputs and components.
  - Repli5 develops advanced vision systems.
  - Binar Solutions contributes expertise in system integration.
  - Hitachi Energy Sweden AB contributes use cases, providing technical inputs and components.
  - Chalmers Industriteknik works with control logics for kitting and advanced vision based and Gen AI-based solutions.

At University West, we conduct research in collaboration with the surrounding world to create a better future.