Synplan: Solving shift scheduling with artificial intelligence

A manager can spend several weeks a year planning a work schedule for healthcare staff that balances the needs of both employees and patients. At the same time, unforeseen events like illness can require quick changes, leading to significant time spent finding replacements on short notice. The Alna district in Oslo has implemented the AI tool Synplan to simplify this process.

"With this tool, employees can be matched to needs when planning shifts. This allows for consideration of both employees' skills and their working hours while also adapting to their preferences," says Leif Smith, HR advisor in Alna district, to Groruddalen newspaper.

"Let’s say an employee is alone with their children for two weeks at a time. They can input that they need day shifts during those weeks but are available for evening shifts during the others. The system also accounts for busy periods like school graduations and the first day of school, providing an early warning when replacements might be needed," says Smith.

The system is based on employees submitting their requests for vacations, days off, and adjustments to their work schedules via questionnaires. Artificial intelligence analyzes the information and generates a shift schedule that meets the need for competence while considering employees' preferences. Managers also receive alerts about periods of increased sick leave so they can bring in extra help.

"We input anonymous data from the healthcare organization we work with, and then it learns a pattern that we humans are unable to see. Based on that, it can predict that next Monday, two people on the day shift will be sick, and one on the evening shift," says Synplan CEO Lars Dahle to NRK.

Synplan, with the help of Innovation Norway, is now set to initiate a collaboration with the Mayo Clinic in Minnesota. Initially, they will explore whether the solution can be integrated into the data system of what is considered the world’s best hospital.

"We are excited to collaborate with forward-thinking minds in the Nordics to uncover new medical knowledge, support healthcare professionals, and improve access to and quality of care," writes Dr. John Halamka, president of Mayo Clinic Platform, to NRK.

"There is probably no better place to have as a reference customer," says Dahle.

Synplan's story began when Trondheim Municipality sought to explore the use of artificial intelligence to predict sick leave and develop better staffing plans. Researchers Hai Nguyen and Pinar Øzturk from NTNU were engaged in a two-year research project in collaboration with the municipality. Nguyen, who has over 12 years of experience working with AI and machine learning-based solutions at Open AI Lab at NTNU and Telenor, is now the CTO of Synplan.

Trondheim Municipality followed up on the research project by becoming one of the first municipalities to implement the system. The results have included reduced overtime, fewer work schedule adjustments, and fewer violations of working time regulations, in addition to lower costs.

"When we started, I thought this might not work. But now I believe this is the future. We've spent significantly less on overtime in 2024, which saves money. And we no longer have to call in people on their days off," says Astrid Lidbom, unit manager at Brundalen Nursing Home, to Adresseavisa.