Two Disruptions That Seem Alike
Last-mile teams often treat same-day rerouting as one basic issue: something breaks, then routes are adjusted. In practice, same-day disruptions have several forms, and logic suited to one can work badly for another.
A driver callout at 7 AM, before departure, is a staffing issue. The stops on that route must be distributed among available drivers before they leave the depot. Key constraints include time windows, vehicle capacity, driver hours, and geographic clustering. It is planning from a fixed starting point.
A missed pickup at 11 AM, three hours into active routes, presents a wholly different challenge. Drivers are spread across the service area. Some have finished stops, others are mid-route, and others are ahead of schedule. The task is not building routes from a fixed starting point. It is fitting the missed pickup into an operation already underway while limiting disruption to committed delivery windows. Constraints include live positions, remaining hours of service, and the opportunity cost of diverting a driver who is on schedule.
These situations need different logic. Treating them as versions of one optimization problem leads many rerouting tools to offer suggestions that work technically but fail operationally.
The Pre-Departure Window
When a driver callout arrives before dispatch, the operation has both its greatest flexibility and its greatest constraints. No one has departed, so the full day's stops can be reassigned. Yet replanning may need to finish within just 15 to 30 minutes, before the first driver leaves.
During that window, a dispatcher must identify remaining drivers with spare capacity, determine which of the called-out driver's stops have hard time windows, and assess the effect on other schedules. For a small fleet, the math is manageable. With 10 to 20 routes and 200 or more stops, the operation needs a tool that can rapidly produce one revised assignment plan across every vehicle.
The pre-departure failure pattern I see is sequential redistribution: move driver A's stops to driver B, test driver B's route, discover it fails, and then try driver C. This loses the broader picture. An effective redistribution may divide the called-out driver's stops among three vehicles in a pattern no single vehicle could absorb, something a one-vehicle-at-a-time process cannot uncover.
The Mid-Route Problem
Mid-route disruptions differ in ways that affect system design. The needed inputs go beyond the remaining stop list. They include each active driver's current location, completed stop count, whether any driver is ahead with slack time, and current traffic on possible diversion routes.
A missed pickup at 11 AM in an industrial area may not call for an immediate dispatch decision. First ask what constraints actually govern coverage. With a hard close time of 3 PM, two or three hours of operating flexibility may remain. If the shipper expects the truck within the hour, the urgency changes.
MatchLog first lays out the constraints surrounding a mid-route disruption, then shows rerouting options. Diverting the nearest driver is useful only when the option also shows the effect on that driver's remaining committed stops. We have seen too many tools offer a lone "nearest driver" choice without revealing that the diversion would make two confirmed delivery windows late. A dispatcher who has experienced that outcome will stop trusting the system.
Why Stop Re-Sequencing Differs From Route Optimization
True route optimization means finding the globally optimal stop sequence for a vehicle under all constraints, which is computationally intensive. Effective algorithms can take seconds to minutes and may produce results that are difficult for a dispatcher to explain to a driver in the field.
Live, same-day re-sequencing is another task. The dispatcher does not require a globally optimal answer. They need a clearly better path than the current trajectory, one the driver can execute with the information in the cab, without replanning stops already completed or underway.
This distinction shapes system design. MatchLog is not a route optimization engine. It offers re-sequencing choices that can be executed under current conditions, a narrower and more useful definition for same-day recovery. Giving a dispatcher a "fully optimized" route that sends a driver back through heavy traffic to revisit a completed delivery cluster is the wrong answer, even if its theoretical distance score is strong.
The Driver Communication Layer
Driver communication is an often overlooked part of last-mile rerouting. When a stop is added, removed, or re-sequenced mid-route, the driver must be informed. The message needs to be clear, immediate, and tied to the driver's current position. Recording a rerouting decision in the system without successfully telling the driver does not resolve the problem.
MatchLog supports the operator's decision-making. Driver communication still uses the operation's chosen channel, whether a direct call, an ELD message, or a push update through a driver app. We are not replacing that layer. Operators who want route changes pushed automatically without dispatcher confirmation are asking for a workflow we do not consider ready, and we say so directly. Driver confirmation remains a safety and compliance checkpoint that automation should not bypass.
What Works in Practice
From what we have observed in early-access operations, rerouting methods that withstand real operating conditions share several traits.
They identify the disruption type before recommending action. A callout, missed pickup, and vehicle breakdown are different events, each needing distinct inputs and choices.
They show effects on other drivers beside the proposed change. Seeing that sending vehicle 7 to the missed pickup will move its final delivery to 6 PM against a 5:30 PM window lets a dispatcher decide. Seeing only the diversion does not.
They let dispatchers confirm a change instead of applying it automatically. In our early-access cohort, dispatchers consistently favored reviewing a suggested re-sequence over automatic driver-assignment updates. That could change as confidence grows, but it is the appropriate default for a tool at this stage.
They keep their scope focused. Tools that claim to reroute same-day work, optimize next-day routes, schedule drivers, and manage customer communication often fall short across all of those jobs. Teams succeeding most with MatchLog usually began with one defined issue, driver callout recovery or mid-route missed pickup handling, and judged us on that issue rather than a broad wish list.