Funding period 01/2020 – 02/2022
The goal of the KI-Para-Mi project is to develop an intelligent personnel planning system for flexible shift scheduling in nursing, which above all takes into account the interests of the employees. The shortage of qualified nursing personnel is a major topic that shapes public debate and political agenda around the globe. Better medical care and rising life expectancy are leading to an increased demand for skilled nursing staff. The gap between demand and actual supply of personnel is growing increasingly. In addition, the average length of stay in the nursing profession is much shorter than in other occupational fields, which places a heavy burden on the body and mind of employees in this profession. Furthermore, classic rigid shift models are still in use, which do not allow flexible shift and duty scheduling. A re-scheduling of already defined shifts is usually only possible with great effort. The inflexible shift plans also make it more difficult to work part time and return to work, e.g. after parental leave, and lead to many trained specialists leaving the profession.
The digital personnel planning system of the project partner Planerio GmbH is to be extended by an AI concept. With the help of AI methods and machine learning algorithms for huge search spaces and ML-based optimization (i.e., CPlex SAT and CPlex Integer Programming, but also the Google API, and especially the Google OR-Tools), the wishes and short-term needs of the employees are to be calculated optimally and more flexibly, based directly on the availability and preferences of users. To facilitate the interaction between the system and its users, the KI-Para-Mi concept features a chat platform based on Rocket.Chat and a chat bot based on the RASA deep learning framework. Both frameworks are state-of-the-art and open source, thereby allowing for a cost efficient deployment.
The consortium is supported by two associated application partners, Lebenswert – Ambulantes Intensivpflegeteam Nordbayern GmbH and Bärenstark Intensivpflege GmbH, with the provision of anonymized real shift schedules. In addition, the evaluation of the KI-Para-Mi demonstrator will be carried out by the associated partners in a field test. The objectives addressed in this project generate a significant added value, both for practice and the economy, as well as for research and society. An increased flexibility of the shift planning system with a focus on the needs of the employees will yield a prolonged duration of stay in the nursing and health profession through increased job motivation and satisfaction of the employees.
Daniel Sonntag (DFKI, supervisor)
Hans-Jürgen Profitlich (DFKI, senior software developer)
Alexander Prange (DFKI, PhD student)
Annika Carolin Grieser (DFKI)