How to calculate truck cycle time without telematics data
Cycle time is the backbone number for haul fleet planning: load, haul, dump, return, queue, repeat. Most of the literature on calculating it assumes you've got a fleet management system feeding GPS pings and load-weigh data straight into a dashboard. That's fine if you're the mine running the fleet. It's no use at all if you're benchmarking a neighboring pit, evaluating an acquisition target, or just don't have telematics access to the operation you need numbers on.
Cycle time can be reconstructed from truck counts and route geometry alone, without a single telematics feed. The method is older than fleet management software. It's just less convenient, and it needs a few assumptions you have to be honest about.
The counting method
The core logic is Little's Law, borrowed from queuing theory: the number of trucks on a route at any moment equals the arrival rate times the average time each truck spends on that route. Rearranged, average cycle time equals the number of trucks observed on the haul network divided by the rate at which trucks are completing cycles (loads per hour, roughly).
To use this without telematics, you need two things measured independently:
- A count of trucks on the haul road network at a given moment, ideally at consistent times across several days.
- An estimate of loads per hour, which you can often back out from stockpile growth, crusher throughput, or public production figures if the site reports them.
Divide one by the other and you get an average cycle time for the fleet on that route. The result is a fleet-level average, the same number most utilization benchmarks run on anyway.
Where this gets practical is the truck count. Ground observation doesn't scale and site access usually isn't available for a competitor benchmark anyway. A daily pass of VHR satellite or aerial imagery at 0.5 m or better resolution can pick out truck positions along the haul loop well enough to count them and roughly bucket where they sit: loaded side, empty side, queued at the crusher, parked at the shovel. Do that at a consistent time of day over a run of days and you've got the arrival-rate input for Little's Law without ever touching the operator's systems.
What road condition tells you that counts can't
Truck count alone tells you how many vehicles are moving. It doesn't tell you why cycle time is drifting if it is. That's where road condition matters as a second read alongside the count.
A haul road that's rutted, wet, or poorly bermed slows haul speed on the loaded leg specifically, which stretches cycle time without changing truck count at all. If you're watching a competitor's fleet and their truck count holds steady but their apparent segment speed (distance over the same stretch, time between two image passes) drops, road condition is the first thing to check before assuming the fleet got bigger or smaller. Visible grading activity, standing water, or washboarding on the haul network is a leading indicator that cycle time is about to move even before the truck count shifts.
This is also the honest limit of the method. You're estimating an average across the fleet from two inputs measured independently, not measuring an individual truck's door-to-door time. Treat the output as a benchmark figure to track over time and compare against your own fleet, not a number precise enough to feed into a single-truck dispatch decision.
Putting it together for a benchmark
For most benchmarking purposes, the number that matters is the trend: whether the competitor's apparent cycle time is getting longer or shorter month over month, and whether that tracks with road condition or a change in truck count. Producing that trend means repeating the read daily, which is the part that's hard to sustain by hand.
A daily count of trucks on the haul network, paired with a read on road condition from the same imagery, is exactly the input this method needs and the thing that's tedious to produce manually day after day. That's the core of what Mine Fleet Activity does for a defined haul network: a daily truck count and road condition read, pulled from imagery alone, built for cycle-time and utilization benchmarks without a site visit.
If you're running this kind of benchmark regularly, it's worth seeing whether a daily imagery-based count can replace the manual version. Have a look at the pilot and see if it fits your haul network.