Miguel Lejeune, a professor of decision sciences and of electrical and computer engineering at the George Washington University, will be visiting UC Davis on Feb. 15 and 16 for a seminar entitled Drone Network Design for Time-Sensitive Medical Events: Queueing MINLP Models, Reformulations, and Algorithms. Join the Davis campus community for the seminar from 11-12 on Wednesday, February 15th in 2310 Gallagher Hall, or sign up to meet one-on-one with Dr. Lejeune during his visit.

His areas of expertise include stochastic programming, large-scale and data-driven optimization, prescriptive and predictive analytics, distributionally robust optimization, decision-dependent uncertainty, financial risk, and supply chain management.  Some of the applications of his recent research projects include vehicle fleet management, mobility-as-a-service for resilience delivery, and improving resilience in the energy sector against wildfires. 

About the Seminar:

Drone Network Design for Time-Sensitive Medical Events: Queueing MINLP Models, Reformulations, and Algorithms

We present queueing-based optimization models to design networks in which the objective is to minimize the response time. The problem is motivated by the drone-based delivery of automated external defibrillators to out-of-hospital cardiac arrests and naloxone to opioid overdoses. The networks are modelled as collections of interdependent M/G/1 or M/G/K queueing systems with fixed and mobile
servers in which the drone service times and the reception of emergency requests are random variables whose distribution parameters are determined endogenously. The optimization models take the form of
nonconvex MINLP problems with fractional and bilinear terms and minimize the average response time which is conducive to maximizing the chance of survival of patients. We derive a reformulation approach
and propose a solution method that features a warm-start component, new optimality-based bound tightening (OBBT) techniques, and an outer approximation algorithm. In particular, we propose new MILP and feasibility OBBT models that can derive multiple variable bounds at once. The computational experiments are based on real-life data from Virginia Beach, and ascertain the computational efficiency of the approach and its impact on the response time and the probability of survival.