Industries / Transport and vehicles / Maritime and shipping / Offshore supply vessels
Scheduling
Scheduling is deciding which supply vessels to hire, which platforms each one visits and when, from which shore base, at what speed, and whether operators share them.
6
Milestones achieved, from 2000 to 2021.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
2000
Supply routes by optimization
Optimization is the branch of math that searches for the cheapest plan that meets every constraint. Kjetil Fagerholt and Håkon Lindstad used it to choose the vessels and weekly schedules for platforms in the Norwegian Sea served from one onshore base, testing six scenarios for platform opening hours and visits per week. The recommended policy showed "potential savings of 7 million dollars," and they report that it was put into use.
Credit
Kjetil Fagerholt; Håkon Lindstad
2009
Slower trips to save fuel
Statoil had the 20 supply ships working for it sail round trips at 12 knots instead of the 15 they could make. On one trip from the Ågotnes base to the Statfjord, Kvitebjørn and Veslefrikk fields that cut more than 6 tonnes of CO2, or 1,900 tonnes a year, by Statoil's figures, and the run to Statfjord took nine hours instead of eight.
Lessonslower sailing saves fuel only when the schedule leaves room for the extra hour.
Credit
Statoil (project manager Ellen Karoline Norlund)
2012
Fleet and routes planned together
Elin Halvorsen-Weare and colleagues published a method that chooses both how many supply vessels to hire and the weekly voyages each one sails. By their account "Statoil has implemented a planning tool based on the voyage-based solution method and reports significant savings," with no figure given.
Credit
Elin E. Halvorsen-Weare, Kjetil Fagerholt, Lars Magne Nonås and Bjørn Egil Asbjørnslett; Statoil
2019
Shared North Sea vessel pools
The logistics company Peterson says it has run vessel sharing in the southern North Sea for more than 20 years and, from about 2017, a pool serving three operators in the central and northern North Sea. It says pooling can cut passage time by up to 50%, worth over £1 million per vessel per year, figures that are its own.
Credit
Peterson
2020
Two operators, one vessel
Wintershall Dea's Brage platform gave up a vessel of its own: the Energy Swan now sails a fixed route that also serves Equinor's Statfjord A, B and C and Veslefrikk platforms, with cargo gathered at one base at Mongstad. Neither company gave a figure for the savings.
Credit
Wintershall Dea Norge; Equinor; Golden Energy Offshore, the vessel's owner
2021
Port work eats sleep
On 25 June 2021 the supply vessel Elliot Cheramie struck an uncrewed platform 77 miles southwest of Port Fourchon, Louisiana, after its mate fell asleep on watch; he had been awake for 19 hours, most of them loading and unloading in port. Four people had minor injuries, and the damage cost $362,814. The NTSB found the owner had not kept to its own 12-hour work limit.
Lessona sailing schedule that ignores the hours spent working cargo in port is not a safe one.
Credit
US National Transportation Safety Board, report MIR-23-01
Watch for
Vessel sharing in one plan
Kabal, a vessel planning software company, says it has worked with Equinor since 2024 on sharing vessels between operators, and that Equinor, Vår Energi and OKEA now plan vessel schedules together "in one unified environment," with a smaller fleet and more frequent service. It gives no figures.
Achieved whenan operator or an independent party publishes measured results, such as vessel days, fuel or CO2 saved.
Claimed by
Kabal, March 2026
Open
01
Measure what sharing saves
A reconstruction, from ship-tracking records, of a year of supply voyages from one base such as Mongstad, showing which vessels sailed, which platforms they called at and for how long, and a plan for the same calls with every vessel shared, to show the vessel days sharing would save.
The briefCredit · unclaimed02
Add weather to test problems
A version of the public test problems for routing platform supply vessels with hourly waves and wind at each platform attached, and a script that scores how a plan holds up when weather stops cargo lifts.
The briefCredit · unclaimed
Open 01 · the brief
Measure what sharing saves
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A reconstruction, from ship-tracking records, of a year of supply voyages from one base such as Mongstad, showing which vessels sailed, which platforms they called at and for how long, and a plan for the same calls with every vessel shared, to show the vessel days sharing would save.
Why it is still open
Sharing deals and software claims come without public numbers: the 2020 Brage agreement gave none, and neither does Kabal's 2026 claim. A SINTEF proof of concept maps supply-vessel tracks on data it does not publish, and an open project that reruns offshore logistics papers on Norwegian maps does not measure real voyages.
Built with what exists
Norway's open AIS data, installation positions from Sodir, the PyVRP or Google OR-Tools route solvers, Python.
First step
Pull one month of AIS tracks for vessels leaving Mongstad and split them into voyages.
Done looks like
A published table of one year of voyages from one base and a shared plan that serves the same calls, with the vessel days of each and the code to rerun both.
Open 02 · the brief
Add weather to test problems
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A version of the public test problems for routing platform supply vessels with hourly waves and wind at each platform attached, and a script that scores how a plan holds up when weather stops cargo lifts.
Why it is still open
The public problem set from Carnegie Mellon University, published with a 2024 paper by Silva, Wang, Filho and Gounaris, gives vessels, orders, fuel costs and time windows for platforms off south-east Brazil, but no weather.
Built with what exists
The Carnegie Mellon problems, Copernicus ERA5 hindcast waves and wind, PyVRP or Google OR-Tools, Python.
First step
Attach ERA5 wave heights to the six platforms of one problem and replay a good plan through one winter.
Done looks like
Every problem re-issued with ten years of hourly weather, and two planning methods scored on late deliveries and extra vessel hours across those years.