Using Genetic Algorithms on a PC to Compile Staff Rosters
Graeme M Snowdon, Phillip Elliott · Journal of Pharmacy Practice and Research · 1998
Aim: We aimed to develop an automated staff rostering system for pharmacy staff working on Saturday rotations that was fair to all concerned, required minimal input and was simple to modify as needs changed. Method: The roster was prepared on a 33 MHz 486 IBM personal computer (PC) in the spreadsheet package Microsoft Excel v4, using Evolver 2.0, the genetic algorithm add‐on for Excel. Results: A staff roster was prepared in the form of a spreadsheet that could be adjusted to meet the constraints of annual leave, rostered days off and many more variables. The genetic algorithm software, linked to the spreadsheet, automatically produced an acceptable roster with no timetable clashes. Conclusion: Pharmacy staff favour the new system since many variables can be assessed with little manual input. A workable roster that is fair and equitable can be easily obtained.