Transport and logistics run on planning, routing, loading and unloading, documentation and communication between links in the chain that rarely see each other. A large part of the work consists of combining fixed rules with changing circumstances: weather, traffic, availability of drivers and equipment, customs rules, customer agreements that differ per assignment. In addition, there is work that revolves around physical actions at a specific location, and work that runs on trust between people who have known each other for years: the planner who knows which driver can handle a difficult route, the forwarder who knows which customer is flexible with a delivery date.
That combination determines where AI can take over work and where it cannot. Planning and routing largely consist of optimization within rules, and that is exactly the type of work systems score well on once the data is correct. Physical actions on location and the judgment call in an exceptional situation remain mostly human work, because something is needed there that does not fit into a form.
Three types of work run through each other in every logistics process. There is work a system can already take over, such as recalculating a route in the event of a road closure or filling in standard documentation based on previous shipments. There is work that is partly taken over, with a human assessing the outcome and approving or rejecting it: a proposed schedule that a planner still reviews before hitting the road, an estimated arrival time that an employee corrects based on knowledge the system does not have. And there is work that remains human work, such as negotiating with a customer about a difficult delivery or assessing a situation on the floor that is not in the data.
In one company, that shift is already well underway: planners work with a system that proposes routes and only escalates the exceptions, freeing up hours for the kind of work that does need people. In another company, everyone still does by hand what a system could already propose, not because the will is lacking but because the data is not in order, the systems do not talk to each other, or no one has established who is allowed to approve a system's proposal and who is not. The difference rarely lies in the technology. It lies in whether the organization has set up the conditions to actually let the work shift.
A board that says "we want AI-driven planning" rarely means the same thing as the planning department that has to execute it. For the board, it is a saving on freed-up hours. For the planning department, it is a question of who remains responsible if a system proposes a route that turned out to be wrong afterward. That difference in interpretation is exactly where the distinction between a goal and an ambition becomes relevant: a goal is measurable and time-bound, an ambition describes a direction that still needs to be translated into what actually changes in the work.
That translation concerns eight readiness dimensions: does the organization have its data in order, are processes described well enough to hand over to a system, is there someone allowed to approve the outcome, is the technology integrated with what is already running, and more. An ambition that already gets stuck at the first gate — because no one knows which data is stored where — is a different ambition from one that gets stuck at the last gate, because decision rights were never established. Both happen in transport, and both call for a different conversation than "we are implementing AI".
That conversation sometimes touches on the question of what happens to existing work when a task disappears. Separate legal requirements apply there, apart from what a readiness assessment can say about capacity.
The tension between what a system can take over and what an organization can bear does not only play out in transport. In professional services, similar questions arise about what AI ambitions are at play in advisory and service work, where the work is less physical but the trade-off between automating and trust returns in comparable form. And within transport itself, another pattern often plays out: the planning department has formulated an ambition that clashes with the ambition of the operations department, and no one named that conflict before the project got stuck. What to do when departmental goals work against each other is described in an approach for clashing departmental goals.
The underlying question is usually not whether AI works in transport, but which work in your own company can actually be transferred and which work remains human work; that is exactly what the FTE TO AI work scan maps out per task. For those who first want to know where their own organization stands, there is the free readiness check: eight short questions, one per dimension, with a picture of where you are furthest along and where you are least far along. The full ambition assessment, with the four layers and the translation into roles and decision rights, is under construction.
Vertel wat u wilt bereiken, dan kijken we samen wat daarvoor moet staan.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.