Many trades in the skilled craft sector are increasingly competing with large companies that can economically manufacture more and more individualised products with the help of automation. At the same time, due to demographic change and unattractive job profiles in the skilled trades, it is becoming more and more difficult to recruit young people.
The use of lightweight robots, which are already being used highly effectively in industry, could significantly increase the competitiveness of the skilled trades for customers and employees. However, the potential of this technology has not yet been fully exploited in small businesses. There is a lack of robotics systems tailored to the needs of the skilled trades that can be widely used at manageable investment costs and that guarantee simple and safe operation.
In this research project, an AI-supported lightweight robot system solution is being developed for cross-sector use in the skilled trades. Craftsmen will receive a low-threshold and intuitively operable system with which they can take control of digital tools and train their robot in craft activities, similar to a master student. The research work focuses, for example, on the robot-assisted sanding and polishing of components.
In order to develop an automated sanding technology, the first step is to analyse sanding by hand or with hand machines in the various trades. Based on this, an end effector for sanding will be developed that combines enhanced sensor technology and intelligent actuator technology.3D scan data and pressure forces will be determined for this purpose, and machine learning will be used to optimize the sanding process and partially automate the robot control. The solutions developed will be prototypically implemented and validated at the participating craft companies.
Simply work with robots
The entire application system is to have greatly improved operational capability, for which learned ‘robot skills’ and intelligent sanding tools can serve. In an automated sanding process, geometries, surface properties and machining paths are to be automatically evaluated and given out by the system without the need for programming individual work steps – up to NO-CODE-ROBOTICS.
This gives the system the character of a flexibly deployable tool that can execute a desired operation in rapid interaction or in an automated manner. In this way, grinding processes could be described via a simple learning teach-in. The users could also let the robot operate independently with movements it has already learned – via so-called machining classes.
A broad applicability of the system to sanding tasks at different trades and company sizes is achieved by the modular design of the software and hardware, as well as a sharing model for sanding with robots.
The use of robotics should thus also be made possible for craftspeople without additional training in the future. As a result, in the medium term, craftspeople could receive a low-threshold and intuitively operable system with which they can independently train robots and implement specific tasks in batch size 1. With the sanding application, craftspeople will be able to take control of digital tools and train their robots in craft activities in a similar way to a master student.
Sanding is a very widespread, everyday process in various trades. Across Germany, for example, around 40,000 carpentry stores, around 3,000 orthotics manufacturers and around 1,200 companies from the musical instrument manufacturing sector can benefit from the solutions. Automation not only enables efficiency gains, but also relieves employees of monotonous, physically demanding work. In addition, it provides an innovative leap forward in the digitization of the skilled trades and offers a professional future perspective for the training and further education of young people. It thus contributes to the preservation and further development of a modern craft tradition.
Craft research of automation solutions
From the craft for the craft
Solutions for the skilled trades are not created in a laboratory, but from concrete observation and need – analysis and prototyping in craft practice are therefore at the beginning and end of every test arrangement.
Designing the change
The knowledge transfer to ‘best practice’ – on the basis of cross-industry, concrete use cases – aims at a high transferability and technological acceptance in the skilled crafts sector.
Research of automation modules
The automation of individual production processes requires intelligent systems. Machine learning helps the skilled trades to digitize their experience and optimize processes.
Prototypical implementation and validation
The ‘proof of concept’ using various tools and processes is the basis for later use – and answers the questions: Which interaction of system and application method is best suited to which task? Or: Can robots be trained in crafts?