An online decision-theoretic pipeline for responder dispatch

Abstract

The problem of dispatching emergency responders to service traffic accidents, fire, distress calls and crimes plagues urban areas across the globe. While such problems have been extensively looked at, most approaches are offline. Such methodologies fail to capture the dynamically changing environments under which critical emergency response occurs, and therefore, fail to be implemented in practice. Any holistic approach towards creating a pipeline for effective emergency response must also look at other challenges that it subsumes -predicting when and where incidents happen and understanding the changing environmental dynamics. We describe a system that collectively deals with all these problems in an online manner, meaning that the models get updated with streaming data sources. We highlight why such an approach is crucial to the effectiveness of emergency response, and present an algorithmic framework that can compute promising actions for a given decision-theoretic model for responder dispatch. We argue that carefully crafted heuristic measures can balance the trade-off between computational time and the quality of solutions achieved and highlight why such an approach is more scalable and tractable than traditional approaches. We also present an online mechanism for incident prediction, as well as an approach based on recurrent neural networks for learning and predicting environmental features that affect responder dispatch. We compare our methodology with prior state-of-the-art and existing dispatch strategies in the field, which show that our approach results in a reduction in response time with a drastic reduction in computational time.

Hierarchical Planning for Dynamic Resource Allocation in Smart and Connected Communities

“…State of the art: DRA and dispatch problems are typically modeled as Markov decision processes (MDP) [22,31,33]. The decision maker's goal is to ind an optimal policy, which is a mapping between system states and actions to be taken.…”

“…There is a broad spectrum of approaches available to address resource allocation under uncertainty as shown in igure 1. In our previous work, we applied the extremes of this continuum to ERM, with each approach having strengths and weaknesses [30,33,36].…”

“…When real world problems are modeled as MDPs, state transitions are typically unknown. The standard method to address this issue is to use a simulator (a model of the world) to estimate an empirical distribution over the state transitions [33], and use the estimates to learn a policy using dynamic programming [22]. Unfortunately, this oline approach does not scale to realistic problem settings [30].…”

“…The standard method to address this issue is to use a simulator (a model of the world) to estimate an empirical distribution over the state transitions [33], and use the estimates to learn a policy using dynamic programming [22]. Unfortunately, this oline approach does not scale to realistic problem settings [30]. Another method is to use an online algorithm like Monte Carlo tree search (MCTS) [30].…”

“…Dynamic resource allocation (DRA) in anticipation of uncertain demand is a common problem in city-scale cyberphysical systems (CPS) [30]. In such a scenario, the decision-maker optimizes the spatial location of resources (typically called agents) to maximize utility over time while satisfying constraints speciic to the problem domain.…”