Fair shift schedules: Saarland University start-up relies on artificial intelligence

Andreas Karrenbauer and Matthias Manderscheid want to reorganize personnel planning by relying on artificial intelligence (from left to right). Credits: Bertram Somieski, Max Planck Institute for Informatics

More and more people want to be able to adapt their working hours to their individual needs and choose their jobs accordingly. Dynamic and flexible schedules could help, but rarely do. The people in charge of these schedules are quickly overwhelmed, and existing computer programs are not helping enough. This should be changed by new software that uses artificial intelligence to learn from employee assessments. A start-up company of Saarland University wants to reorganize personnel planning and is supported by the IT incubator of Saarland University and the Max Planck Society.

“The complexity increases exponentially and is already noticeable when only ten employees, their wishes and working hours have to be recorded over 30 days. A person can no longer cope with this,” explains Andreas Karrenbauer. He advises the three founders of the start-up “ChronoFair”, who are supported by the Federal Ministry of Economics and Energy with the EXIST start-up grant. Karrenbauer works at the Max Planck Institute for Computer Science on the Saarland Informatics Campus in Saarbrücken. There, he is researching how complex problems can be solved using computational methods.

Together with Matthias Manderscheid, Karrenbauer designed the basics of the online system “ChronoFair”. Not only does ChronoFair calculate the respective shift schedule according to given constraints, it also enables employees to evaluate the schedule online. The system then processes the feedback received in this way with the help of machine learning in order to deliver even better personnel plans tailored to individual needs in the future. “So far, scheduling has been too difficult due to the communication between the participants, not to mention the computational complexity,” explains Karrenbauer.

ChronoFair combines web technology and artificial intelligence: First, the planner enters all necessary data on employees and shift systems online. Employees can also communicate their wishes and preferences via a browser. The system summarizes all this data and forms a complex optimization problem, which, depending on the number of employees and the shift system, can comprise several million equations and inequalities. From the possible solutions, ChronoFair selects the best one by means of a special evaluation function and proposes it as a work schedule. The planner and employee can now evaluate individual shifts and entire shift sequences online. “With this feedback and machine learning, we can calculate nuances that do not play a role in the operational process, but make the difference between shift paradise and work hell for employees,” says Karrenbauer. The founders have just completed the final software tests in cooperation with Hydac. The Saarland company allows ChronoFair to make its calculations based on the shift models and personnel concepts used in the company without passing on personnel data. The Saarländische Wagnisfinanzierungsgesellschaft is already investing in the start-up, whose founders also include Paul Manderscheid and Sven Foit.

Further information:
www.chronofair.de

Questions can be directed to:
Andreas Karrenbauer
Max Planck Institute for Computer Science
Saarland Informatics Campus E1.4
Phone: +49 681 9325 1007
E-mail: karrenba@mpi-inf.mpg.de

Matthias Manderscheid
ChronoFair
Saarbruecker Str. 3c
66901 Schoenenberg-Kuebelberg
E-mail: m.manderscheid@ChronoFair.de
Mobile: +49 170 3700981

Editor:
Gordon Bolduan
Competence Center Computer Science Saarland
Saarland University
Saarland Informatics Campus E1.7
Phone: +49 681 302-70741
Email: bolduan@mmci.uni-saarland.de