In online freight bidding, carriers or brokers face a closed-loop stochastic decision problem where they must accept or reject tendered loads in real-time, considering operational feasibility, fleet repositioning costs, and opportunity costs against future demand. Public, reproducible benchmarks for this problem are scarce: existing routing benchmarks are static, while dynamic fleet studies often rely on private operator data.
We introduce FreightBidBench, a public-calibrated, dependency-free, closed-loop benchmark where feasibility (pickup reach, appointment windows, simplified hours-of-service, stochastic yard delays) and economics (service-failure penalty, terminal fleet value, daily price-premium window) are explicit, versioned, and reproducible from public Freight Analysis Framework and U.S. Department of Agriculture truck rate data.
We develop two full-horizon hindsight ceilings: a simple LP-style relaxation and a tighter Lagrangian-per-truck information relaxation, which retains per-truck hours-of-service and sequencing structure, being 20.7% tighter than the LP relaxation in a tight-capacity scenario and 39.3% tighter in a scarce-capacity scenario.
We introduce a parametric surrogate-rollout cascade with boundary-band and scarcity-pressure escalation triggers. In ten-seed tight and scarce scenarios, the best simple policy retains 91.0% and 86.5% of rollout profit, while the standard-library surrogate retains 94.2% and 89.3%; a cascade at a single escalation band recovers roughly 98% on both at 40-56% of rollout's mean decision latency, and in the tight scenario is statistically indistinguishable from the rollout teacher (paired-bootstrap 95% CI on the profit delta spans zero).
Blogger's Review: This paper introduces FreightBidBench, providing a significant public benchmark for online freight bidding and addressing the lack of dynamic data in previous studies. With efficient algorithms and relaxation techniques, it enhances the feasibility and economics of decision-making, advancing research in the logistics field. This study offers new insights and frameworks for future dynamic optimization.