Everyone who buys online knows the messages telling them when the parcel will be delivered. Nothing wrong with that. But there is a huge variation in the promised delivery times. I live in Belgium and there the promise of today's delivery is predominant. Today means somewhere between 8 am and 8 pm. This is of course a very wide time window. In the Netherlands, time windows are usually narrower, so it is easier for the recipient to arrange for someone to be home.
Customer satisfaction or lower costs?
Timelines in themselves are nevertheless an important service qualifier and a major contributor to customer satisfaction. As a logistics service provider, it is therefore important to deliver within the time window as best as possible. From a cost perspective for the LDV, it is of course more convenient to plan without time windows. Then the planner has more flexibility to schedule the orders as well as possible and the costs are lower. Fixed time windows imposed (by the customer or otherwise) give less scheduling flexibility and higher costs. This is the (academic) basic difference of scheduling with or without time windows.
A double benefit
The above example of parcel delivery is actually more nuanced in reality. The logistics service provider informs the recipient of the time of delivery. In doing so, it sets itself a time slot and communicates this to the customer. But in fact, the logistics service provider is trying to achieve a double benefit here: low costs in organising its routes (without many restrictions) and keeping the recipient satisfied by providing a (relatively narrow) delivery time slot. In their planning, they can effectively treat it as if there were no time slots at all. This means full flexibility and low costs. They then communicate a time slot to the recipient based on their planned routes. The customer is satisfied too.
Now, however, it is no longer possible to adjust routes. If the customer is not at home after all, or there is a traffic jam, or... the route may not be adhered to after all. Then it is necessary to be able to re-plan quickly and effectively, taking into account the already announced time slots of the original route. Unfortunately, re-planning is an expensive hobby: no flexibility (e.g. time slots are fixed now) and done under time pressure. It also always leads to less efficient routes.
Scheduling uncertainty
Research shows that if you plan better in advance, taking into account the potential uncertainty (travel times, customers, etc.), you will ultimately be better off as a logistics provider. The planning itself becomes more complicated, albeit supported by smart ICT. But ultimately, less to no re-planning is needed because the routes being driven are already robust to these uncertain events. This arguably leads to lower costs, a higher probability of always being able to deliver in time windows and much less real-time correction of routes.
Tom van Woensel - Professor of Freight Transport & Logistics
