TRTal Raviv

Tal Raviv

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Tal Raviv has a total of 64 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Routing in Line Networks with Handling Times

    by Gabriel Deza, Michal Tzur, Tal Raviv

    Motivated by hyperconnected urban parcel networks for same-day and time-sensitive delivery, we study the fixed-line routing problem that arises when evaluating a candidate set of recurring interhub vehicle services. Vehicles follow predefined cyclic sequences of service points, or lines, and parcels may transfer between lines. Such transfers expand path choice but require loading and unloading operations that make vehicles dwell at service points. Because handled volumes are routing-dependent, dwell times become endogenous, linking routing decisions to line cycle times, effective service frequencies, and transfer waiting times. We introduce the routing in line networks with handling times problem and formulate it as a bilinear generalized capacity-flow assignment model. We characterize optimal dwell times and the resulting frequencies as functions of handling workloads, enabling a reformulation solely in terms of commodity flows. We exploit its convex substructure to derive two second order conic relaxations that preserve the original feasible region and, therefore, yield feasible incumbent solutions along with valid lower bounds. We also develop a residual-frontier framework that characterizes the trade-off between the convex relaxation objective and the remaining nonconvex residual, yielding a general mechanism for constructing lower bound certificates. Computational experiments show that ignoring handling times can yield poor or infeasible routing, whereas accounting for endogenous dwell times can improve average parcel travel times by up to 26.5%. In instances with up to 257 nodes and 25 lines, the relaxations solve quickly, yield near-optimal routing, and obtain small optimality gaps, whereas off-the-shelf spatial branch-and-bound struggles to produce useful bounds. Funding: This research was supported by the Israel Science Foundation [Grant 1460/20]; The first author was partially supported by the Ministry of Aliyah and Integration, the Shlomo Shmeltzer Institute for Smart Transportation and the Israeli Smart Transportation Research Center. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2026.0039 .

  • Open Access

    The lecture-hall quadratic assignment problem

    by Tal Raviv

    We study the post-timetabling \emph{lecture-hall assignment problem}, in which a fixed weekly timetable and realized student enrollments are given, and the decision is to assign each lecture to a compatible hall so as to minimize the walking burden imposed on students moving between consecutive lectures and a penalty for feasible but non-ideal compatibility of halls-lecture. The problem combines a quadratic assignment objective, driven by the product of shared enrollments and hall-to-hall distances, with interval-scheduling feasibility constraints and hall-capacity restrictions. We prove that the problem is NP-hard and that every approximation lower bound for the classical quadratic assignment problem carries over, already in a single-day, two-block special case. We develop three exact formulations: a bilinear mixed-integer quadratic program, a compact mixed-integer linear program with one auxiliary distance variable per successor pair, and a constraint-programming model. All of these formulations admitting a natural daily decomposition. To tighten the relaxations we introduce biclique distance cuts that dominate the standard pairwise linking inequalities, together with Chv\'atal-Gomory capacity-dominance cuts and a safe compatibility-preprocessing procedure. The framework also accommodates common timetabling side constraints such as hard and soft \emph{SameRoom} and \emph{SameAttendees} requirements, for which biclique-strengthened linearizations are derived. A comprehensive numerical experiment covers daily instances derived from five large ITC~2019 competition benchmarks and from a 2023 institutional dataset from Lancaster University. The compact linearized program emerges as the most robust formulation, and biclique strengthening is the dominant modeling device: it closes virtually all of the root-relaxation gap and lifts the number of proven-optimal solves substantially, while cutting median solver runtimes by roughly an order of magnitude. The results demonstrate that realistically sized lecture-hall assignment instances can be solved to optimality or near-optimality within minutes, and frequently even within seconds, by off-the-shelf mixed-integer programming solver once the biclique structure of the quadratic walking term is exploited.

  • DynoPath: A Dynamic Online Grid-Based Centralized Sorting Algorithm

    by Ilan Estrugo, Yossi Bukchin, Tal Raviv

    High-throughput sorting facilities require significant investment in costly resources, including labor, space, and equipment. The rapid growth of e-commerce coupled with economic and technological advancements in recent decades has intensified the demand for sorting technologies with increased throughput and improved cost/performance ratio. In this paper, we investigate sorting systems that utilize grids of four-way conveyors, where items can move freely in up to four cardinal directions from each array cell to its neighbors. In such systems, items enter the grid at designated input cells, and they are routed in a sequence of steps to their target output cells and eventually, to their destination bins. The grid-based approach enables multiple items to move simultaneously, offering the potential for significantly higher throughput compared with traditional conveyor-based sorting systems. However, operating grid-based sorting systems requires solving complex, real-time, parallel decision-making challenges. A naive, myopic operational policy is susceptible to possible deadlocks, where items block each other from reaching their destinations. To address this, we propose a novel centralized online algorithm that ensures deadlock-free operation. The algorithm models item movements within the grid using a time-expanded graph. Upon an item’s arrival, its route and schedule are determined, and the graph is immediately updated to preclude conflicts with future items. Extensive experiments demonstrate that our centralized algorithm achieves higher throughput than previously proposed distributed algorithms. Moreover, it scales effectively, enabling real-time operation of large grids with thousands of cells, which makes it suitable for modern high-throughput sorting facilities. Funding: This research was supported by the Israel Science Foundation [Grant 2236/21]. I. Estrugo was supported by a scholarship from the Shlomo Shmeltzer Institute for Smart Transportation at Tel-Aviv University. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2025.0033 .

  • Open Access

    Unsupervised Classification under Uncertainty: The Distance-Based Algorithm

    by Alaa Ghanaiem, Evgeny Kagan, Parteek Kumar, Tal Raviv, Peter Glynn, Irad Ben-Gal

    This paper presents a method for unsupervised classification of entities by a group of agents with unknown domains and levels of expertise. In contrast to the existing methods based on majority voting (“wisdom of the crowd”) and their extensions by expectation-maximization procedures, the suggested method first determines the levels of the agents’ expertise and then weights their opinions by their expertise level. In particular, we assume that agents will have relatively closer classifications in their field of expertise. Therefore, the expert agents are recognized by using a weighted Hamming distance between their classifications, and then the final classification of the group is determined from the agents’ classifications by expectation-maximization techniques, with preference to the recognized experts. The algorithm was verified and tested on simulated and real-world datasets and benchmarked against known existing algorithms. We show that such a method reduces incorrect classifications and effectively solves the problem of unsupervised collaborative classification under uncertainty, while outperforming other known methods.


  • Optimal Retrieval in Puzzle-Based Storage Systems Using Automated Mobile Robots

    by Tal Raviv, Yossi Bukchin, René de Koster

    Puzzle-based storage (PBS) systems store unit loads at very high density, without consuming space for transport aisles. In such systems, each load is stored on a moving device (conveyor module or transport vehicle), making these systems very expensive to build and maintain. This paper studies a new type of PBS system where loads are moved by a small number of autonomous mobile robots (AMRs). The AMRs (or vehicles) can travel freely underneath loads and lift a specific load and carry it to a neighboring vacant space. These systems are hard to analyze, as all the AMRs can move simultaneously with or without loads. We formulate an integer linear programming model that minimizes the retrieval time and the number of load and vehicle movements. The proposed model can handle single-load movements as well as block movements, multiple input/output points, and various constraints on simultaneous vehicle movements. The integer linear programming formulation can solve relatively small problems (a grid with up to about 50 cells) and a sufficient number of empty cells. For larger systems or those with few empty cells, a three-phase heuristic (3PH) is developed, which significantly outperforms the heuristic methods known to date and solves large instances sufficiently fast. The 3PH and an additional hybrid heuristic yield relatively small gaps from a lower bound provided by the integer linear programming model. We find that increasing the number of vehicles has a diminishing return effect on the retrieval times. Using a relatively small number of vehicles makes retrieval times only slightly longer than those obtained when having a vehicle under each load (which is equivalent to the traditional PBS systems). With single-load movement, more vehicles are needed compared with block movement to reach short retrieval times. Also, the marginal contribution of extra empty slots appears to decrease rapidly, which implies high storage densities can be obtained in practice.