The AI-PRISM project is oriented towards a human-centred AI-based solutions ecosystem targeted to manufacturing scenarios with tasks difficult to automate and where speed and versatility are essential. To facilitate the assessment of the performance, transferability, scalability, and large-scale deployment of AI-PRISM solutions, we will conduct research under real operational environments in five pilots involving key manufacturing sectors: furniture, food/beverage, built-in appliances, Electronics and one generic discrete manufacturing.
In light of the furniture manufacturing demonstrator, Project Leader entity NTT DATA Spain and researchers from the Polytechnical University of Valencia (ai2 and CIGIP), Fondazione Bruno Kessler (FBK), and the University of Cranfield (CRAN) had the opportunity to participate in an exciting data collection experiment that combined RGBD cameras and eye-tracking technology to digitise the art of sanding in the facilities of partner Andreu World (AW) in Valencia, Spain. This partner uses cutting-edge technology to industrialise the traditional craftsmanship works of manufacturing furniture. This project perfectly exemplifies the intersection between traditional craftsmanship and modern technology. By leveraging AI-based data-driven techniques, we can improve upon age-old techniques and bring them into the digital age.
The data collection experience to digitise the art of sanding
During the visit, using RGBD cameras and eye tracking technology, the AI-PRISM team captured the movements and postures of the specialists as they sanded a piece of furniture while also tracking their subjective view and generating heatmaps of the areas of interest. Researchers used these innovative tools and captured the essence of the movements and postures of specialists as they sanded a piece of furniture while tracking their subjective view and generating heatmaps of the areas of interest.
The pilot aims to create digital models of the specialist, the piece of furniture, and the surrounding work environment. These models will support the human-centric design and development of innovative, collaborative robotic solutions. In addition, the aim is to improve the sanding process’s performance and ergonomics while preserving the artisanal skill that makes this craft so unique, as we can teach robotic agents to mimic the artisanal skills of human workers.
Challenges and next steps in the pilot
The main technical challenges are linked to the complex perception abilities and dexterity that are put into play on this task. Using their hands, specialists feel the wood grain and the texture of the finished furniture parts and apply the right place to achieve a perfect result. Other challenges identified by the researchers during their visit were:
- many occlusions which prevent a complete reconstruction of both the human worker and the chair he is operating with;
- sawdust settles on chair parts during processing preventing a clean acquisition of the wood surface;
- understanding the sequence of actions that a human worker performs to sand a chair part, a (recurrent) neural network should also consider their temporal dependence over a long-time span;
- the human worker does not rely on his sight only, but his touch plays an important role in detecting wood imperfections.
The AI-PRISM team is excited to see where this project will take us. We will apply the same methods and technologies in the other industries that conform to our five pilot use cases, requiring high manual dexterity and expertise.
Do you want to learn more about AI-PRISM and the team behind the project? Subscribe to our newsletter and follow us on LinkedIn and Twitter, so you don’t miss a thing!