Skip to ContentSkip to Navigation
Centre for Operational Excellence (COPE)Part of University of Groningen
Centre for Operational Excellence (COPE)
Faculty of Economics and Business
Centre for Operational Excellence (COPE) Projects Easicon - Efficiency and Sustainability in Construction

Planning for delays without buying for the worst case: fleet sizing for a construction hub when project start dates shift.

Renovation work in the Eemsdelta region of Groningen, where buildings damaged by induced earthquakes are being restored, depends on a steady flow of materials to many small sites. A construction hub can bundle those materials and deliver them just in time, but only with the right vehicles. Buying too many ties up money in trucks that stand idle, while buying too few forces expensive last-minute outsourcing. The difficulty is that the hub must choose its fleet months ahead, while the start dates of the projects it serves keep moving. A project delayed by permits or staff shortages shifts all of its deliveries with it.

The EaSiCon project develops a digital twin that gives a real-time virtual representation of construction processes and their connected supply chains, supporting tactical and operational planning to improve efficiency and sustainability. This article presents research into strategic fleet sizing for a construction hub, which examines how many and which vehicles are needed to serve renovation sites on time when project start dates are uncertain. The study combines a long-term fleet decision with day-to-day delivery planning, tested on a case based on the Eemsdelta renovation programme.

Just-in-time delivery to dispersed sites

For every renovation, a daily material kit is prepared at the hub, holding everything the crew needs for that day's work. Electric trucks deliver the kits and return to the hub each evening to recharge. Timing is strict. A kit that arrives on the day of work is too late and stalls the crew, while a kit delivered too early has nowhere to go, because sites have little storage space and residents live nearby. Unlike a large building site receiving bulk loads, the hub serves many scattered addresses with a few pallets each.

Planning under many possible futures

Rather than planning for one schedule, the study generates a large number of possible futures. In each, every project starts on time, early, or late, with delays more likely than early starts. The model first chooses a fleet and then works out the best delivery plan for each of those futures, weighing vehicle investment against transport costs.

Planning every individual route over twelve weeks and hundreds of futures would take far too long to compute. The study therefore uses a simplified model that decides on which days each site is served and estimates the driving distance, without drawing the exact routes. Checked against full route planning, this shortcut proposes almost the same delivery days, bundles deliveries in the same way and comes very close in cost, while producing a fleet recommendation within minutes.

A compact fleet covers almost every case

For the Eemsdelta workload, the model consistently arrives at a small mixed fleet of one small and one big truck. That fleet delivers every kit on time in all but a very small share of the futures tested. Covering those last extreme cases as well would require swapping the small truck for a second big truck, a noticeably higher investment that buys protection against delays that rarely occur. Where a hub can hire in extra transport on the odd bad day, the compact fleet is the better choice.

The analysis also shows where capacity actually runs short. Driving range is never the limiting factor, since the vehicles use well under their full range on a typical day. Pallet space is the real constraint. Many trucks leave fully loaded, while others run partly empty on days when strict delivery windows leave little room to combine sites.

Conclusion and recommendations

Uncertain project timing does not have to mean a large, expensive fleet. By planning against many possible schedules at once, a hub can find the smallest fleet that still delivers reliably and make a conscious choice about how much rare risk it is willing to carry. Practitioners setting up a hub should:

  • size the fleet on realistic schedule deviations, not on the nominal plan
  • treat pallet capacity, not driving range, as the constraint to plan around
  • weigh the cost of covering rare extreme delays against occasional outsourcing
  • use a simplified planning model for long horizons and detailed routing for daily operations

The Eemsdelta workload is modest, so the next step is to apply the same approach to larger construction programmes and to add uncertainty in material volumes alongside uncertainty in timing. A construction digital twin can supply exactly those up-to-date schedules and volumes, so that fleet decisions keep pace with the projects they serve.

For more information, please contact:

Ragnar Klabbers

Ivo Agricola

Kees Jan Roodbergen

Michael Dienstknecht

Jeroen Laarman

Last modified:21 October 2026 5.07 p.m.