In most factories, dispatch is work that begins where production ends but has no full counterpart in any system. When an order is ready, someone picks up the phone, calls the hauliers they know, asks for a price and arranges a vehicle. Which job went to whom and for how much sits in an Excel file, sometimes only in that person's notebook. When the vehicle arrives the waybill is written by hand, the goods are loaded, and from that moment the shipment becomes invisible; when the customer calls to ask, the answer is again found by telephone, by ringing the driver. At the end of the year, the answer to how much was paid for transport is the sum of the invoices in accounting. Nobody can break down how much transport cost was carried by which customer, which route and which product. Yet transport is a visible line in the unit cost of most products, and when a price is quoted it is glossed over with a guess.
In this area two separate jobs are often called by the same name and get confused. Transport management, that is TMS (Transportation Management System), manages the haulage itself: which orders will be consolidated onto the same vehicle, which carrier will take which job at which freight rate, when the vehicle will be loaded, on which day the goods will be delivered. Fleet management, by contrast, is about running your own vehicles: fuel consumption, maintenance and tyres, inspection and insurance dates, driver licences and cost per vehicle. A company without vehicles of its own does not need fleet management but does need a TMS; a company that distributes with its own fleet needs both. We design these two sides as a single system, but keep them separated so that they can be rolled out independently.
The layer we build does not replace your existing ERP; it begins where that ends. Order, stock and invoice records stay in Logo, Mikro, Netsis (Turkish ERP and accounting products) or whichever system you use. The TMS layer reads those orders, turns them into dispatch orders, produces the vehicle and route plan, assigns the job to a carrier and writes the delivery result back to its source. On the driver's side there is a mobile screen: loading, delivery, signature, photograph and the reason for a failed delivery are entered there. That record lands in the system as proof of delivery, that is POD (Proof of Delivery). So whether the goods were delivered is answered from a screen rather than by telephone, and in disputed deliveries you hold a dated, signed, photographed record. Picking and packing inside the warehouse sits outside this; the TMS manages what happens outside the gate.
Expectations need to be set in the right place from the start. Route optimisation is the most overhyped heading in this field: the system can produce a mathematically good sequence, but it cannot fully model traffic, a vehicle waiting at the gate, the customer's goods-in hours or the driver's knowledge of the ground. We therefore set optimisation up as a suggestion rather than an instruction; we ask the planner to intervene in the plan and to record the reason for intervening. The second reality is address data. In Türkiye a significant share of customer addresses is held as free text and does not land on the right point on the map; address cleansing is the invisible but longest-running part of the work in most projects. The third is acceptance on the driver's side: a driver who does not use the mobile screen can render the whole system useless on his own.