This particular colleague lends a hand whenever there is a problem with a machine during the production process— in the event of disruptions, downtimes, capacity being overloaded or rejects being produced. In these situations, they quickly find out what is causing the fault interrupting production. In a matter of seconds, they evaluate all the data at their disposal—all the manuals, information and experience they have been fed by their colleagues. Their name is MADDOX, and they are an assistance system, programmed on AI-basis, for manufacturing and processing machinery used in the packaging industry.
MADDOX will step in whenever a disruption occurs. The algorithm then assesses whether this or a comparable fault situation has occurred before and quickly develops solutions to solve the problem for the machine operators’ benefit. This is particularly helpful for new, inexperienced colleagues, says Andre Schult, the CEO of Peerox, the firm that developed MADDOX. But by no means is it only rookies who stand to benefit: “The assistance system ensures that ramp-up times are accelerated, productivity is increased by reducing downtimes, and the skills shortage in the manufacturing sector is addressed,” says the 40-year-old, who has a mechanical engineering background. He and his business partner Markus Windisch founded the company in Dresden five years ago. Since then, knowledge management has been in high demand. There are now 19 members of staff working at the start-up, whose head office is located between Dresden University of Technology (TU Dresden) and software specialist Robotron.