Optimizing operations management in uncertain times
When variables are in constant flux, when the stakes are high and times seem uncertain, operations management becomes a tricky beast to master. Research teams at the HEC Montréal Department of Logistics and Operations Management work to develop increasingly sophisticated decision-support tools based on mathematical optimization and artificial intelligence.
Minimizing ambulance response times in situations where every minute counts. Ensuring that essential businesses have the food people need in stock, even when consumers are rushing to the shelves in a panic. They may seem unrelated, but these logistics challenges are rooted in the same science.
When skillfully harnessed, data and artificial intelligence can guide decision-making in uncertain or unpredictable contexts.
Stepping up in an emergency
Valérie Bélanger is an associate professor at HEC Montréal. Her research focuses on the logistics of emergency medical transportation, a complex environment that is constantly evolving and where every decision has high stakes. “In densely populated urban areas, ambulances make frequent but short trips,” she explains. “In rural areas, there are fewer emergency calls, but first responders have to travel much greater distances. Sometimes, patients even need to be evacuated by helicopter.” The research team’s challenge was to develop logistical models for both of these scenarios, which were sometimes completely different from one another.
Driven by the conviction that university research has a strong potential for practical applications, aside from its scientific importance, Bélanger and her team worked with Urgences-santé to design a simulation tool that allowed the paramedic organization to test various scenarios before implementing changes in the field. What are the most strategic places to position vehicles on standby? What hospitals should patients be taken to? How many ambulances are needed to cover a given area? These questions can be answered using the decision-support tools developed by Bélanger and her team.
Every year, over 680,000 ambulance trips are completed in Quebec. Optimizing their logistics has a direct impact on the cost and efficiency of the ambulance network.
Retail businesses: Responding to demand
For a researcher like Yossiri Adulyasak, holder of the Canada Research Chair in Supply Chain Analytics, who focuses on the disruption of supply chains, a global pandemic is an unparalleled opportunity to
develop and test concrete solutions. The HEC Montréal professor and his team wasted no time in getting down to business. Their research lab? Canadian retail businesses.
Buffeted by fluctuations in demand for specific product categories (toilet paper, flour, disinfectants, etc.), these businesses have found themselves facing true challenges with inventory management. In response, researchers used big-data analysis techniques and machine learning to help them improve supply chain efficiency and maintain a high service level.
“In a context as uncertain as a pandemic of this magnitude, there are some things you simply will not be able to plan for,” says Adulyasak. “But with machine learning and data analytics on our side, we can detect anomalies in consumer behaviour and plan inventory levels accordingly.” That’s the philosophy behind these decision-support tools: Giving retailers the ability to react in real time. Once alerted to these anomalies in buying patterns, a business can limit the amount of food that can be purchased or accurately predict the optimal time to restock goods. And better planning means better cost management.
For these two professors, this is just the beginning. Valérie Bélanger and Yossiri Adulyasak continue to work with their partners not only to refine the tools they have developed, but also to develop new ones—always with the ultimate goal of providing better service to the public.