# Martin SavelsberghGeorgia Institute of Technology | GT · School of Industrial and Systems Engineering

Martin Savelsbergh

Ph.D.

## About

274

Publications

114,743

Reads

**How we measure 'reads'**

A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more

19,620

Citations

Additional affiliations

November 2011 - present

**University of Newcastle**

November 2011 - July 2014

July 2010 - October 2011

Education

August 1983 - May 1988

September 1978 - May 1983

## Publications

Publications (274)

With the recent boom of the gig economy, urban delivery systems have experienced substantial demand growth. In such systems, orders are delivered to customers from local distribution points respecting a delivery time promise. An important example is a restaurant meal delivery system, where delivery times are expected to be minutes after an order is...

To achieve logistic efficiency and customer convenience in last-mile delivery processes, a system with alternative delivery points in the form of locker box stations can be used. In such a system, customers can be served either at their home address within a certain time window, or at a locker box station where parcels can be picked up at any time....

We explore the benefits of multivariable branching schemes for linear-programming-based branch-and-bound algorithms for the 0-1 knapsack problem—that is, the benefits of branching on sets of variables rather than on a single variable (the current default in integer-programming solvers). We present examples where multivariable branching has advantag...

Less-than-truckload carriers rely on the consolidation of freight from multiple shippers to achieve economies of scale. Collected freight is routed through a number of transfer terminals at each of which shipments are grouped together for the next leg of their journeys. We study the service network design problem confronted by these carriers. This...

We introduce an effective and efficient iterative algorithm for solving the continuous-time service network design problem. The algorithm achieves its efficiency by carefully and dynamically refining partially time-expanded network models so that only a small number of small integer programs, defined over these networks, need to be solved. An exten...

Waiting at the right location at the right time can be critically important in certain variants of time-dependent shortest path problems. We investigate the computational complexity of time-dependent shortest path problems in which there is either a penalty on waiting or a limit on the total time spent waiting at a given subset of the nodes. We sho...

The Hunter Valley coal export supply chain in New South Wales, Australia, is of great importance to the Australian economy. Effectively managing its logistics, however, is challenging, because it is a complex system, covering a large geographic area and comprising a rail network, three coal terminals, and a port, and has many stakeholders, e.g., mi...

Authors are encouraged to submit new papers to INFORMS journals by means of a style file template, which includes the journal title. However, use of a template does not certify that the paper has been accepted for publication in the named journal. INFORMS journal templates are for the exclusive purpose of submitting to an INFORMS journal and should...

We investigate substitution-based equipment balancing for a package express carrier operating multiple equipment types in its service network. The weekly schedule of movements used to transport packages through the service network leads to changes in equipment inventory at the facilities in the network. We seek to reduce this change, i.e., the equi...

Split delivery routing problems are concerned with serving the demand of a set of customers with a fleet of capacitated vehicles at minimum cost, where a customer can be served by more than one vehicle if beneficial. They generalize traditional variants of routing problems and have applications in commercial as well as humanitarian logistics. Previ...

We explore the benefits of multi-variable branching strategies for linear programming based branch and bound algorithms for the 0-1 knapsack problem, i.e., of branching on sets of variables rather than on a single variable (the current default in integer programming solvers). We present examples where multi-variable branching shows advantage over s...

Integer programs for solving time-dependent models—models in which decisions have to be made about the times at which activities occur and/or resources are utilized—are pervasive in industry, but are notoriously difficult to solve. In the last few years, interest in the role of discretization in approaches to solve these problems has intensified. O...

This work describes a genetic algorithm based approach for the optimization of the Hunter Valley coal export system in Newcastle, Australia. The Port of Newcastle features three coal export terminals, operating primarily in cargo assembly mode. They share a rail network on their inbound operations and a channel on their outbound operations. Maximiz...

We present a new solution approach for the Time Dependent Traveling Salesman Problem with Time Windows. This problem considers a salesman who departs from his home, has to visit a number of cities within a predetermined period of time, and then returns home. The problem allows for travel times that can depend on the time of departure. We consider t...

We present the theoretical foundations for a number of preprocessing and cut generation techniques for multi-objective binary programs. The techniques are based on a characterization of conditions under which the objective functions of a multi-objective binary program guarantee the existence of an ideal point in criterion space, i.e., the existence...

We consider a setting in which a rail infrastructure provider divides the track costs proportionally between the above-rail operators based on their usage. We study a proposed access charge regime aimed at incentivizing the operators to use longer train configurations. The regime sets a target length and gives a discount on an operator’s charge if...

The Port of Newcastle features three coal export terminals, operating primarily in cargo assembly mode, that share a rail network on their inbound side, and a channel on their outbound side. Maximising throughput at a single coal terminal, taking into account its layout, its equipment, and its operating policies, is already challenging, but maximis...

We present an exact algorithm for the Minimum Duration Time-Dependent Shortest Path Problem with piecewise linear arc travel time functions. The algorithm iteratively refines a time-expanded network model, which allows for the computation of a lower and an upper bound, until - in a finite number of iterations - an optimal solution is obtained.

Less-than-Truckload (LTL) transportation carriers plan for their next operating season by deciding: (1) a load plan, which specifies how shipments are routed through the terminal network from origins to destinations, and (2) how many trailers to operate between each pair of terminals in the network. Most carriers also require that the load plan is...

We study a class of problems wherein a vehicle departs from an initial location, often
called a depot, visits each location in a known set of locations exactly once and during a
pre-defined time window, and then returns to the depot. The problem of determining a
set of such movements, what is commonly referred to as a tour, is known as the Travelin...

We study the connection between biobjective mixed integer linear programming and normal form games with two players. We first investigate computing Nash equilibria of normal form games with two players using single-objective mixed integer linear programming. Then, we define the concept of efficient (Pareto optimal) Nash equilibria. This concept is...

Researchers and practitioners have long recognized that many transportation problems can be naturally and conveniently modeled using time-expanded networks. In such models, nodes represent locations at distinct points in time and arcs represent possible actions, e.g., moving from one location to another at a particular point of time, or staying in...

Different solution methods are developed to solve an inventory routing problem for a perishable product with stochastic demands. The solution methods are empirically compared in terms of average profit, service level, and actual freshness. The benefits of explicitly considering demand uncertainty are quantified. The computational study highlights t...

The success of on-demand service platforms, e.g., Uber and Lyft to obtain a ride and Grubhub and Eat24 to get a meal, which rely on crowd-sourced transportation capacity, has radically changed the view on the potential and benefits of crowd-sourced transportation and delivery. Many retail stores, for example, are examining the pros and cons of intr...

Seamless integration of ride-sharing and public transit may offer fast, reliable, and affordable transfer to and from transit stations in suburban areas thereby enhancing mobility of residents. We investigate the potential benefits of such a system, as well as the ride-matching technology required to support it, by means of an extensive computation...

We present a linear programming based algorithm for a class of optimization problems with a multi-linear objective function and affine constraints. This class of optimization problems has only one objective function, but it can also be viewed as a class of multi-objective optimization problems by decomposing its objective function. The proposed alg...

Consolidation carriers transport shipments that are small relative to trailer capacity. To be cost effective, the carrier must consolidate shipments, which requires coordinating their paths in both space and time; i.e., the carrier must solve a service network design problem. Most service network design models rely on discretization of time-i.e., i...

We propose the vehicle routing problem with roaming delivery locations (VRPRDL) to model an innovation in last-mile delivery where a customer’s order is delivered to the trunk of his car. We develop construction and improvement heuristics for the VRPRDL based on two problem-specific techniques: (1) efficiently optimizing the delivery locations for...

The Traveling Salesman Problem with Time Windows is the problem of finding a minimum-cost path visiting each of a set of cities exactly once, where each city must be visited within a specified time window. It has received significant attention because it occurs as a subproblem in many real-life routing and scheduling problems. We explore an approac...

We study the vehicle routing problem with roaming delivery locations in which the goal is to find a least-cost set of delivery routes for a fleet of capacitated vehicles and in which a customer order has to be delivered to the trunk of the customer’s car during the time that the car is parked at one of the locations in the (known) customer’s travel...

We introduce Adaptive Kernel Search (AKS), a heuristic framework for the solution of (general) Mixed Integer linear Programs (MIPs). AKS extends and enhances Kernel Search, a heuristic framework that has been shown to produce high-quality solutions for a number of specific (combinatorial) optimization problems in a short amount of time. AKS solves...

We introduce Adaptive Kernel Search (AKS), a heuristic framework for the solution of (general) Mixed Integer linear Programs (MIPs). AKS extends and enhances Kernel Search, a heuristic framework that has been shown to produce high-quality solutions for a number of specific (combinatorial) optimization problems in a short amount of time. AKS solves...

We study a variant of an abstract scheduling problem inspired by the management of reclaimers in the stockyard of a coal export terminal. We prove NP-completeness of the problem and formulate it as a mixed-integer program. We show that for a given reclaiming sequence, the problem can be solved in pseudo-polynomial time. In addition, we provide simp...

The multi-depot vehicle routing problem (MDVRP) is a variant of the standard vehicle routing problem in which there are multiple depots at which vehicles can start and end their routes. The authors develop an integer programming-based heuristic for the MDVRP. In the authors’ heuristic, the authors use integer programming to improve solutions to the...

We propose a proactive route guidance approach that integrates a system perspective:
eliminating congestion, and a user perspective: minimizing distance traveled.
The approach assigns paths to users so as to minimize congestion while not
increasing their distance traveled too much. By limiting the set of paths considered
for an origin-destination p...

In the Generalized Independent Set Problem, we are given a graph, a revenue for each vertex, and a set of removable edges with associated removal costs, and we seek to find an independent set that maximizes the net benefit, i.e., the difference between the revenues collected for the vertices in the independent set and the costs incurred for any rem...

The desire of companies to offer same-day delivery leads to interesting new routing optimization problems. We study the complexity of single depot dispatching problems in which a delivery to a customer must occur within a pre-specified time after the customer places the order. Thus, each order has a release date (when the order is ready for dispatc...

Less-Than-Truckload (LTL) carriers operate networks of consolidation terminals, and route each customer shipment through a sequence of transfer terminals. At each terminal stop, a shipment is unloaded from an inbound trailer and reloaded onto an outbound trailer. A load plan determines the specific path of terminals to be used to transfer freight m...

We conduct an extensive computational study to quantify the impact of different types of participants’ flexibility on the performance of a single-driver, single-rider ride-sharing system. Our results consistently show that small increases in flexibility, e.g., in terms of desired departure time or maximum detour time, can significantly increase the...

The Traveling Salesman Problem with Time Windows (TSPTW) is the problem of finding a minimum-cost path that visits each of a set of cities exactly once, where each city must be visited within a given time window. The TSPTW has received substantial attention, over many years, as it is core to many important, real-life routing and scheduling applicat...

Stochastic programming models are typically solved through Benders decomposition, a "dual solution method" in which a sequence of infeasible (super-optimal) solutions is generated until a first feasible (in fact, optimal) solution is eventually found and the method terminates. As convergence may require a large amount of computing time for hard ins...

We consider a setting in which a company not only has a fleet of capacitated vehicles and drivers available to make deliveries, but may also use the services of occasional drivers who are willing to make a single delivery using their own vehicle in return for a small compensation if the delivery location is not too far from their own destination. T...

We present a new algorithm for optimizing a linear function over the set of efficient solutions of a multiobjective integer program (MOIP). The algorithm’s success relies on the efficiency of a new algorithm for enumerating the nondominated points of a MOIP, which is the result of employing a novel criterion space decomposition scheme which (1) lim...

We present a new variant of the full 2-split algorithm, the Quadrant Shrinking Method (QSM), for finding all nondominated points of a tri-objective integer program. The algorithm is easy to implement and solves at most single-objective integer programs when computing the nondominated frontier, where is the set of all nondominated points. A computat...

We investigate two routing problems that arise when order pickers traverse an aisle in a warehouse. The routing problems can be viewed as Euclidean travelling salesman problems with points on two parallel lines. We show that if the order picker traverses only a section of the aisle and then returns, then an optimal solution can be found in linear t...

We study a rail track access charging policy proposed by the Australian Rail Track Corporation (ARTC), in which a discount on access charges is offered if above-rail operators employ the “efficient train”. The efficient train is a train with a particular length, which results in the efficient use of a train path. The ARTC uses train paths to alloca...

For static traffic assignment problems, it is well-known that (1) the total travel
time in a user-equilibrium solution can be substantially higher than the total travel
time in a system-optimum solution, and (2) the user-experienced travel time in
a system-optimum solution can be substantially higher that the user-experienced
travel time in a user-...

Seamless integration of ride-sharing and public transit may offer fast, reliable, and affordable transfer to and from transit stations in suburban areas thereby enhancing mobility of residents. We investigate the potential benefits of such a system, as well as the ride-matching technology required to support it, by means of an extensive computation...

We investigate the potential benefits of introducing meeting points in a ride-sharing system. With meeting points, riders can be picked up and dropped off either at their origin and destination or at a meeting point that is within a certain distance from their origin or destination. The increased flexibility results in additional feasible matches b...

In the mixed capacitated general routing problem, one seeks to determine a minimum cost set of vehicle routes serving segments of a mixed network consisting of nodes, edges, and arcs. We study a bi-objective variant of the problem, in which, in addition to seeking a set of routes of low cost, one simultaneously seeks a set of routes in which the wo...

We present the first criterion space search algorithm, the triangle splitting method, for finding all nondominated points of a biobjective mixed integer program. The algorithm is relatively easy to implement and converges quickly to the complete set of nondominated points. The algorithm maintains, at any point in time, a diverse set of nondominated...

We present a new criterion space search algorithm, the balanced box method, for finding all nondominated points of a biobjective integer program. The method extends the box algorithm, is easy to implement, and converges quickly to the complete set of nondominated points. Because the method maintains, at any point in time, a diverse set of nondomina...

We investigate two routing problems that arise when order pickers traverse an aisle in a warehouse. The routing problems can be viewed as Euclidean travelling salesman problems with points on two parallel lines. We show that if the order picker traverses only a section of the aisle and then returns, then an optimal solution can be found in linear t...

We present a new criterion space search method, the L-shape search method, for finding all nondominated points of a triobjective integer program. The method is easy to implement, and is more efficient than existing methods. Moreover, it is intrinsically well-suited for producing high quality approximate nondominated frontiers early in the search pr...

Disease management programs have emerged as a cost-effective approach to
treat chronic diseases. Appointment adherence is critical to the success of
such programs; missed appointment are costly, resulting in reduced resource
utilization and worsening of patients' health states. The time of an
appointment is one of the factors that impacts adherence...

Growing congestion is a problem faced by cities around the world. Traditional solutions considered include new capital works to increase network capacity and expanding public transport offerings to make public transport more attractive. Dynamic ridesharing is a recent alternative in which people with similar travel plans are matched and travel toge...

We consider a real-world cargo assembly planning problem arising in a coal
supply chain. The cargoes are built on the stockyard at a port terminal from
coal delivered by trains. Then the cargoes are loaded onto vessels. Only a
limited number of arriving vessels is known in advance. The goal is to minimize
the average delay time of the vessels over...

We study simultaneous transportation procurement auctions from a truckload carrier’s perspective.We formulate a stochastic bid price optimization model aimed at maximizing the carrier’s expected profit. The model accounts for synergies among lanes and competing carriers’ bid patterns. We develop an iterative coordinate search algorithm to find high...

The NP-hard single machine sequence-dependent group scheduling problem with minimization of total weighted earliness and tardiness is investigated. An arc-time-indexed formulation is presented and a Lagrangian-based branch-and-bound algorithm is proposed. The Lagrangian relaxation of the arc-time-indexed formulation is solved as a shortest path pro...

Strategic capacity planning is a core activity for the Hunter Valley Coal Chain Coordinator as demand for coal is expected to double in the next decade. Optimization and simulation models are used to suggest and evaluate infrastructure expansions and operating policy changes. These models require input data in the form of shipping stems, which are...

We investigate the potential benefits of introducing meeting points in a ride-sharing system. With meeting points, riders can be picked up and dropped off either at their origin and destination or at a meeting point that is within a certain distance from their origin or destination. The increased flexibility results in additional feasible matches b...

We introduce a class of incremental network design problems focused on investigating the optimal choice and timing of network expansions. We concentrate on an incremental network design problem with shortest paths. We investigate structural properties of optimal solutions, show that the simplest variant is NP-hard, analyze the worst-case performanc...

The need for viable public transit systems has been well documented and so has the role that so-called flexible transport systems can play. Flexible transport services offer great potential for increases in mobility and convenience and decreases in travel times and operating costs. One such service is the demand responsive connector, which transpor...

The Feasibility Pump (FP) has proved to be an effective method for finding feasible solutions to mixed integer programming problems. FP iterates between a rounding procedure and a projection procedure, which together provide a sequence of points alternating between LP feasible but fractional solutions, and integer but LP relaxed infeasi-ble solutio...

We present the first criterion space search algorithm, the triangle splitting method, for finding the efficient frontier of a biobjective mixed integer program. The algorithm is relatively easy to implement and converges quickly to the complete set of nondominated points. A computational study demonstrates the efficacy of the triangle splitting met...

Transport broadly, and freight transport specifically, whether by land, by water, or by air, impacts the environment. Understanding and quantifying the impacts of transport on the environment has been an ongoing pursuit for many years. For example, the report “Transport in a fast changing Europe” of the working Group “Transport 2000+” set up by the...

We consider a real-world cargo assembly planning problem arising in a coal supply chain. The cargoes are built on the stockyard at a port terminal from coal delivered by trains. Then the cargoes are loaded onto vessels. Only a limited number of arriving vessels is known in advance. The goal is to minimize the average delay time of the vessels over...

Caring for patients with chronic illnesses is costly—75% of U.S. healthcare spending can be attributed to treating chronic conditions (CDC, 2009a,b)15.
CDC. (2009b) At a glance 2009: chronic disease - the power to prevent, the call to control. Technical report, Centers for Disease Control and Prevention, National Center for Chronic Disease Preventi...

We consider the variant of the shortest path problem in which a given set of paths is forbidden to occur as a subpath in an optimal path. We establish that the most-efficient algorithm for its solution, a dynamic programming algorithm, has polynomial time complexity; it had previously been conjectured that the algorithm has pseudo-polynomial time c...

This paper presents PACE, an optimisation-based Possession Assessment and Capacity Evaluation decision support tool that efficiently and effectively evaluates schedules of planned maintenance and renewal work for rail infrastructure, allowing users to quickly and comprehensively assess the work schedules against the competing objectives of asset re...

We study a number of variants of an abstract scheduling problem inspired by
the scheduling of reclaimers in the stockyard of a coal export terminal. We
analyze the complexity of each of the variants, providing complexity proofs for
some and polynomial algorithms for others. For one, especially interesting
variant, we also develop a constant factor...

A critical component in a coal export supply chain is the stockyard at a terminal, as this is where coal from different mines is “mixed” in a stockpile before being loaded onto vessels. We develop a tree search algorithm for managing the stockyard aimed at maximizing the throughput. The algorithm relies on the analysis and use of geometric properti...

We consider a setting in which a not-for-pro�t service provider divides the total costs of providing the service proportionally among firms based on their usage of the service. We study a new mechanism for incentivizing firms to manage their externalities by setting a threshold for these externalities, and, conditioned on reaching the threshold, gi...

We study an incremental network design problem, where in each time period of
the planning horizon an arc can be added to the network and a maximum flow
problem is solved, and where the objective is to maximize the cumulative flow
over the entire planning horizon. After presenting two mixed integer
programming (MIP) formulations for this NP-complete...