Artificial Bee Colony Optimization to Reallocate Personnel to Tasks Improving Workplace Safety
Worldwide, just under 5,800 people go to work every day and do not return because they die on the job. The groundbreaking Industry 4.0 paradigm includes innovative approaches to improve the safety in the workplace, but Small and Medium Enterprises (SMEs) – which represent 99% of the companies in the EU – are often unprepared to the high costs for safety. A cost-effective way to improve the level of safety in SMEs may be to just reassign employees to tasks, and assign hazardous tasks to the more cautious employees. This paper presents a multi-objective approach to reallocate the personnel of a company to the tasks in order to maximize the workplace safety, while minimizing the cost, and the time to learn the new tasks assigned. Pareto-optimal reallocations are first generated using the Non-dominated Sorting artificial Bee Colony (NSBC) algorithm, and the best one is then selected using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). The approach was tested in two SMEs with 11 and 25 employees, respectively.
KeywordsBee colony algorithm Occupational safety and health Multi-objective optimization Personnel reallocation Risk perception TOPSIS
This research was supported by the PRA 2016 project “Analysis of Sensory Data: from Traditional Sensors to Social Sensors”, funded by the University of Pisa.
- 1.International Labour Organization Statistics Database (ILOSTAT). http://www.ilo.org
- 2.Bandini, S., Manzoni, S., Sartori, F.: Case-based reasoning to support work and learning in small and medium enterprises. In: 21st IEEE International Conference on Tools with Artificial Intelligence, pp. 253–260 (2009)Google Scholar
- 6.Cococcioni, M., Lazzerini, B., Pistolesi, F.: A semi-supervised learning-aided evolutionary approach to occupational safety improvement. In: IEEE Congress on Evolutionary Computation (2016)Google Scholar
- 9.Health, Safety Executive: Reducing error and influencing behaviour. HSE Books (1999)Google Scholar
- 13.Lazzerini, B., Pistolesi, F.: Classifying workers into risk sensibility profiles: a neural network approach. In: 8th European Modelling Symposium, pp. 33–38 (2014)Google Scholar
- 17.Rakshit, P., Sadhu, A.K., Bhattacharjee, P., Konar, A., Janarthanan, R.: Multi-robot box-pushing using non-dominated sorting bee colony optimization algorithm. In: Panigrahi, B.K., Suganthan, P.N., Das, S., Satapathy, S.C. (eds.) SEMCCO 2011. LNCS, vol. 7076, pp. 601–609. Springer, Heidelberg (2011). https://doi.org/10.1007/978-3-642-27172-4_71CrossRefGoogle Scholar