The digital transformation is underway – and, surprisingly, it is racing right past HR departments for the most part. New job descriptions and training requirements are posing major challenges: How do you find the right skills to drive the digital transformation? And above all, how can your own workforce be trained and equipped for network systems? There is no shortage of offerings in the training market, and it is hard to get an overview of them. The crucial part is matching, said Ronald Orth, who directs the Business Excellence Methods Department at the Fraunhofer Institute for Production Systems and Design Technology (IPK): How do you decide which employee is best suited for which new role? What training do they need to acquire the knowledge necessary for the work? What expertise has to be built up and expanded? And how do you find the right learning opportunities in the rather opaque education market? The answer: with artificial intelligence.

Researchers have been teaming up with industrial partners in the “KIRA Pro” project, developing a system of the same name. To help with the selection of training programs, KIRA Pro takes into account employees’ preferences and learning strategies. The consortium partner Peers Solutions oversees the development of the technology tools and has already brought out its learning path generator Selena. The startup was spun off from the technology company Trumpf in 2019. So there was originally an emphasis on production, development and quality-management roles. The firm has now become active in other industries and disciplines. Philipp Ramspott, Chief Learning Officer at Peers Solutions, spells out how the AI-supported tool works based on the job of a quality manager working with battery cell production. It is a prime example since a predominantly manual activity of the past is being transformed by data mining and machine learning. As graduate mechanical engineers, quality managers certainly have basic analytical knowledge, Ramspott said. But so far they have lacked skills in evaluating and interpreting large amounts of data.

At this point, companies and their skilled workers are posing three questions: How do you translate trends into specific skill requirements? How do you choose courses in the heavily fragmented education market? And how do I guarantee the transfer of what I learn into daily life? The answer is the “role navigator,” an automated system that casts a line into the sea of education offerings, hoping to haul in courses that match the company‘s goals and the employee‘s profile. It is not a job that a hiring manager can do. “The goal is not to develop another platform full of training courses,” Orth said. "Instead, the idea is to take existing offerings that are available on different platforms, bundle them with a special purpose in mind, and process them into a personalized learning path." That's how matching a course with the (training) needs of the employees works. It is easier if the roles that are adaptable to the company's specific situation are defined in advance. Orth uses an example from manufacturing to explain how the tool works: Based on its transformation strategy, an automaker decides that the role of the multimachine operator will become more important in the future. Available employees with appropriate profiles are to be retrained for the work. The current qualifications of the employee in question are fed into the role navigator. It then checks to see which training courses could lead to new skills matching the target role, based on the employees' knowledge and abilities. "In this way, a learning path is developed for each employee with various custom-fit training modules, which collectively would prepare him for the target role,” he said. The tool decides which employees will most likely help the company reach its goals: For example, a programmer would have to complete a long learning path to qualify as a multimachine operator. On the other hand, mechatronics technicians could be trained for the role on a straight-line learning path with far fewer stages.