In freight dispatching, a familiar paradox persists: a route plan built at 6 a.m. rarely survives the realities of 9 a.m. traffic, driver illness, or a carrier that goes offline. HERE Technologies is addressing this challenge not by building a better static plan, but by developing a system that continuously learns from what happens once that plan leaves the dispatcher's office.

Bart Coppelmans of HERE Technologies outlined the company's product roadmap in an interview with FreightWaves at Home Delivery World. The discussion covered upgrades to HERE's tour planning engine, a newly launched driver feedback tool called Last Meter Guidance, and a prototype AI route optimization reasoning layer designed to explain its decisions rather than simply issue them. The focus reflects a broader operational problem for logistics teams: route optimization only creates value if dispatchers and drivers can adapt when field conditions diverge from the plan.

The Planning-Execution Gap

Tour planning ranks among HERE's most established services, with a development history spanning a decade. Recent feature additions are driving increased market adoption, according to Coppelmans.

"We started developing this ten years ago, but it's really picking up in the market now as one of the best performing solvers, especially because of what we added last year," Coppelmans said.

The most significant of those additions is time-dependent optimization, which factors in how fluctuating traffic conditions affect delivery capacity throughout the day.

"At nine o'clock in the morning you can deliver fewer orders than at one o'clock in the afternoon because of traffic jams," Coppelmans said.

HERE also introduced driver-friendly overlapping tours, which reduce the territory conflicts that frustrate drivers on their routes. Additionally, a walk clustering feature detects when a driver should park once and complete several deliveries on foot instead of repeatedly entering and exiting a vehicle.

"From one parking spot you can then deliver by walking to multiple different deliveries in a certain area, which might be more efficient than driving in and out of your vehicle," Coppelmans said.

Last Meter Guidance Closes the Loop

These planning improvements do not resolve what Coppelmans identified as the core challenge: the disconnect between what dispatch plans in the morning and what a driver actually faces in the field.

"If you have a perfect plan by six in the morning, by nine it can already be different because of unexpected events — a driver getting sick, a carrier going dark or last-minute order changes," Coppelmans said. "You need to be really dynamic and flexible, taking that into account."

HERE's response is Last Meter Guidance, a client-side service that operates on a handheld device or driver application and gathers sensor and positioning data from the field. In dense delivery environments, the final approach to a stop can be materially different from simply navigating to a street address, especially when parking, building entrances and walking paths determine how long a delivery actually takes.

"We're automatically collecting sensor and probe positioning points, we have our own positioning stack," Coppelmans said. "This service builds on top of that, really making sure we're learning from the field. We're collecting traces data from where the vehicle is parking, the walk path toward the building, flagging the building entrance and the final delivery end-point location."

That data flows bidirectionally. Dispatchers receive more precise delivery windows, while drivers gain parking and entrance guidance based on where previous drivers successfully completed deliveries, rather than relying solely on where a map places a building's entrance.

"There's no disconnect anymore," Coppelmans said. "Drivers are more comfortable trusting what is being planned and can say, 'Okay, this makes sense.'"

AI Route Optimization Learns to Explain Itself

Layered on top of both services is what HERE describes as a route optimization cognitive layer — a prototype agentic capability that the company expects to enter closed beta later in 2026. While the underlying tour planning API instructs dispatchers on what actions to take, the reasoning layer is designed to explain why those decisions were made. That distinction matters in dispatch operations because a recommendation that appears inefficient may be tied to constraints the user cannot immediately see.

"Why are these orders unassigned? Why are these two trucks going down the same street on the same day?" Coppelmans said. "It might be because of actual constraints, driving skills, or certain priorities."

The layer extends beyond explanation. It is also built to propose corrective actions — the types of adjustments an experienced dispatcher might make intuitively.

"Maybe loosen certain constraints, move some orders to tomorrow, or add two vehicles into the capacity," Coppelmans said. "That's the domain-specific reasoning layer."

Coppelmans drew a parallel to developments already underway in telematics, where fleets deploy generative AI to investigate why a tire is losing pressure or why an asset has gone missing. Applied to dispatch operations, the same methodology means a manager no longer needs to independently deduce every downstream cause of a service failure.

"It gives you proactive responses in terms of what you can do to further improve your plan and make it even better," Coppelmans said.

Grounding AI Before It Hallucinates

Alongside the reasoning layer, HERE announced a separate location reasoning layer engineered to prevent large language models from fabricating information when queried about geography.

When asked to find a restaurant halfway along a truck route, generic LLMs frequently produce unreliable results, Coppelmans noted.

"If you now ask generic LLMs about a certain geo location, you get really random results, totally off, wrong geometry or wrong location," Coppelmans said. "You get a wrong POI that isn't nearby the river but somewhere else. You can get fooled easily, and these LLMs hallucinate based on geo-location queries."

HERE Location Reasoning is designed to provide AI agents with the spatial grounding necessary to answer such questions accurately — whether that involves understanding a vehicle's position on a road, drawing a boundary around a midpoint, or filtering for points of interest that genuinely lie along a given route. For transportation applications, that grounding is central because geographic errors can affect routing, stop selection and the instructions passed to drivers.

"That's the journey we're on, further supporting different agents being built in the market," Coppelmans said. "We're feeding them with a correspondent layer so they can really understand the context of location and ground it."

The Agent-to-Agent Future

Coppelmans does not envision the reasoning layer replacing a company's own operational judgment. Each carrier's KPIs and service-level agreements vary too widely for a single shared model to address independently, he noted.

"You need to have your own agentic operations agent running, but feeding that with learnings from others is super crucial, otherwise you're siloed," Coppelmans said.

The longer-term outlook, he suggested, resembles less a single company owning one AI model and more a network of agents querying one another — a carrier's dispatch agent consulting a routing agent much as a person might seek a colleague's second opinion. For now, the nearer-term marker to watch is whether the closed beta can show that explainable route recommendations help dispatchers manage exceptions without losing control over operating rules.

"You might think more in terms of agent-to-agent communications connecting certain things," Coppelmans said. "It doesn't necessarily have to be fully integrated into your own system, you can call different agents to pull other data sets to verify and qualify. But that's not something we're at yet, it's a little bit further down the road."