Integrating Artificial Neural Networks, Simulation and Optimisation Techniques in Ambulance Deployment for Heterogeneous Regions under Stochastic Environment

Tichaona Wilbert Mapuwei^1, , Oliver Bodhlyera² and Henry Mwambi²

¹Department of Statistics and Mathematics, Bindura University of Science Education, Bindura, Zimbabwe

²School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Pietermaritzburg, South Africa

Cite this paper:
Tichaona Wilbert Mapuwei, Oliver Bodhlyera and Henry Mwambi. Integrating Artificial Neural Networks, Simulation and Optimisation Techniques in Ambulance Deployment for Heterogeneous Regions under Stochastic Environment. American Journal of Applied Mathematics and Statistics. 2022; 10(3):80-94. doi: 10.12691/ajams-10-3-3

Abstract

The paper focuses on the development of a strategy to integrate forecasting using artificial neural networks (ANN), simulation and optimisation techniques for ambulance deployment to predefined locations with heterogeneous demand patterns under stochastic environments. The metropolitan city of Bulawayo was used as a case study with high variability in call inter-arrival rates, response times, service times, and proportions of severity of emergencies by geographical zones covered by sub-stations. These stochastic environments complicate the decision-making process at strategic, tactical and operational level, in pursuit to achieve high levels of equality, efficiency and effectiveness in resource allocation and utilisation. This paper proposes an integrated simulation optimisation methodology that integrates future demand and allows for simultaneous evaluation of operational performances of deployment plans using multiple performance indicators such as average response time, total duration of a call-in system, number of calls in response queue, average queuing time, throughput ratios and ambulance utilisation levels. Increasing the number of ambulances influences the average response time below a certain threshold. Beyond this threshold, no significant changes occur in the performance measures. As the fleet size is increased, the ambulance utilisation levels decreased, hence there is always need to balance resource allocation and capacity utilisation to avoid idleness of essential equipment and human resources. Numerical experiments conducted to align the response time to international standards resulted in reduction in number of ambulances required for optimal deployment. For medical resources such as ambulances, deploying more resources do not always translate to better performance, hence there is need to simultaneously consider multiple performance measures. Decision makers in EMS must seriously consider ways of reducing the response time as it has significant bearing in reducing the required number of ambulances, a critical but scarce resource. Efforts must be directed towards digitisation of switch boards in the call center, training of the paramedics and provision of relevant modern equipment to the response teams as it will go a long way in reducing the pre-trip delay time, chute time and ultimately the response time. Based on the scientific evidence, management could lobby for de-congestion and resurfacing of old and dilapidated roads in order to increase access and speed when responding to emergency calls.

Keywords:
forecasting artificial neural networks simulation optimisation ambulance deployment

This work is licensed under a Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Figures

References:

[1]	Sayed, M.J. (2012). Measuring quality in emergency medical services: a review of clinical performance indicators. Emergency Medicine International, 2012.
[2]	Zaffar, M.A., Rajagopalan, H.K., Saydam, C., Mayorga, M., and Sharer, E. (2016). Coverage, survivability or response time: A comparative study of performance statistics in ambulance location models via simulation-optimisation. Operations Research for Health Care, 11, 1-12.
[3]	Aartun, H.A., Andersson, E.S., Christiansen, H., Granberg, M., and Anderson, T. (2017). Strategic ambulance location for heterogeneous regions. European Journal of Operational Research, 260(1), 122-133.
[4]	Zhen, L., Wang, K., Hongtao, H., and Daofang, C. (2014). A simulation optimization framework for ambulance deployment and relocation problems. Computers and Industrial Engineering, 72, 12-23.
[5]	Aringhieri, R., Bruni, M.E., Khodaparasti, S., and Van Essen, J.T. (2017). Emergency medical services and beyond: Addressing new challenges through a wide literature review. Computers and Operations Research, 78, 349-368.
[6]	Kitapci, O., Ozekiciglu, H., Kaynar, O., and Tastan, S. (2014). The effect of economic policies applied in Turkey to the sale of automobiles: multiple regression and neural network analysis. Social and Behavioral Sciences, 148, 653-661.
[7]	Rather, A.M., Agarwal, A., and Sastry, V.N. (2015). Recurrent neural network and a hybrid model for prediction of stock returns. Expert Systems with Applications, 42, 3234-3241.
[8]	Hastie, T., Tibshirani, R., and Friedman, J. (2008). The Elements of statistical learning: Data mining, inference and prediction. (2nd ed.). New York: Springer Science + Business Media. (Chapter 11).
[9]	Mitrea, C.A., Lee, C.K.M., and Wu, Z. (2009). A comparison between neural networks and traditional forecasting methods: A case study. International Journal of Engineering Business Management, 1(2), 19-24.
[10]	Kheirkhah, A., Azadeh, A., Saberi, M., Azaron, H., and Shakouri, H. (2013). Improved estimation of electricity demand function by using of artificial neural network, principal component analysis and data envelopment analysis. Computers and Industrial Engineering, 64, 425-441.
[11]	Belanger, V., Ruiz, A., and Soriano, P. (2019). Recent optimization models and trends in location, relocation, and dispatching of emergency medical vehicles. European Journal of Operations Research, 272(1), 1-23.
[12]	Zhang, Z.H, and Li, K. (2015). A novel probabilistic formulation for locating and sizing emergency medical service stations. Annals of Operations Research, 229(1), 813-835.
[13]	Boujemaa, R., Jebali, A., Hammami, S., Ruiz, A., and Bouchriha, H. (2018). A stochastic approach for designing two-tiered emergency medical systems. Flexible Services and Manufacturing Journal, 30(1-2), 123-152.
[14]	Erkut, E., Ingolfsson, A., and Erdogan, G. (2008). Ambulance location for maximum survival. Naval Research Logistics (NRL), 55(1), 42-58.
[15]	Handerson, S. and Mason, A. (2005). Ambulance planning: Simulation and data visualization in: Brandeau M.I., Saintfort F., Pierskalla w.p. (eds) Operations Research and Health Care: A Handbook of Methods and Applications, pages 77-102. Kluwer, Academic, Boston, 2004.
[16]	Eldabi, T., and Young, T. (2007). Towards a framework for healthcare simulation. In 2007 Winter, Simulation Conference, pages 1454-1460. IEEE.
[17]	Silva and P.M.S., Pinto, L.R. (2010). Emergency medical systems analysis by simulation and optimization. In Proceedings of the 2010 winter simulation conference, pages 2422-2432. IEEE.
[18]	Pinto, L., Silva, P., and Young, T. (2015). A generic method to develop simulation models for ambulance systems. Simulation and Modelling Practice and Theory, 51, 170-183.
[19]	Aboueljinane, L., Sahin, E., Jemai, Z., and Marty, J. (2014). A simulation study to improve the performance of an emergency medical service: Application to the French Val-de-Marne department. Simulation Modelling Practice and Theory, 47, 46-59.
[20]	Ingolfsson, A., Budge, S., and Erkut, E. (2008). Optimal ambulance location with random delays and travel times. Health Care Management Science, 11(3), 262-274.