Guides for av fleet operations managers
AV fleet availability: how to report cars in service, charging, in maintenance and out of service
"52 of 60 available" means nothing until everyone agrees what available is. One status per car, three buckets, the roster behind the rate, and the hours a car is down split by whether anyone is working on it.
By XoomPark Editorial Team ·
| Status | Bucket | The car is | Borrowed from |
|---|---|---|---|
| In service | Available | On the road, carrying or ready to carry a rider | Mission capable: can perform at least one mission3 |
| Staged | Available | Released and parked, waiting for demand | Spares stay in the transit count4 |
| Charging | Turning | On a charger, not yet at its release level | Our assumption |
| Cleaning or inspection | Turning | In its turnaround clean or walk-around | Our assumption |
| Maintenance, being worked | Down | In a bay with a technician on it | Not capable because of unit level maintenance1 |
| Maintenance, waiting on parts | Down | Stopped because a part is not there | Work stoppage due to a supply shortage1 |
| Held out of service | Down | Kept back by a safety, recall or policy hold | Our reason codes guide |
| Away | Down | At a vendor, dealer or body shop | Depot not available time1 |
| Unaccounted | Down | On the roster, but nobody can say where or why | Possessed but not reported1 |
What counts as an available car?
The services also measure a stricter rate, fully mission capable, for systems that can perform all of their assigned missions.3 A depot meets the same fork with a car that works but has a limit, such as one that may only run short trips until a part arrives. Decide once whether a limited car counts as available, write the rule down, and report limited cars as their own line if you count them. This rule is ours, an assumption and not a standard.
One caution from the vocabulary: availability can be affected by delays before any maintenance and support action is recognised, for example fault detection time and administrative delay.2 A car with a fault nobody has logged still shows as available. The rate is only as honest as the speed at which a status gets changed.
Which cars belong in the roster count?
The same database keeps a separate list of inactive vehicles: those in storage, emergency contingency vehicles, vehicles pulled from the active fleet but awaiting sale, and vehicles out of service for an extended period of time for major repairs.4 That last group is where a rate gets flattered. Our rule, an assumption: a car moves to inactive only by a dated decision with a name on it, never by drifting off the board.
Say which roster the rate is on. The Defense Department keeps two: materiel availability measures the total active inventory, and operational availability measures the primary mission inventory.1 Its own worked example has 691 vehicles in total and 600 assigned to missions. The first roster was available for 226,994 of 252,215 possible days, or 90.0 percent; the second for 200,000 of 219,000, or 91.3 percent.1 Same vehicles, same year, two honest numbers. The gap is the back-up inventory, which exists to permit maintenance without reduction of the assets available for assigned mission requirements.1
Charging and cleaning: the bucket between up and down
Turning is the bucket a schedule can shrink. Maintenance hours depend on what broke; charging hours depend on when cars plug in, how fast the stalls are and how many there are. When stalls are what hold the turning share up, XoomPark arranges AV fleet charging at existing stations in hours their stalls sit unused, never blocks a station's regular customers and never pushes a station past its monthly peak.
A car moves from turning to available at one gate: the readiness release. Until it passes, it is still turning, whatever the charger says.
Working or waiting: how to split the hours a car is down
- The Defense Department classifies not available time, the time an asset is incapable of performing its mission safely, into four categories: maintenance, supply, depot work, and possessed but not reported.1
- The maintenance clock starts, for unscheduled work, when a malfunction is discovered.1
- The supply clock starts when work stops for lack of parts and the parts are not available for at least one hour. Maintenance time resumes when the items are delivered.1
- Depot time starts when the depot inducts the asset.1 For a car, that is the moment the vendor or body shop takes it.
- Any possible time not accounted for as available or not available is counted as possessed but not reported.1 The unexplained hours get a line of their own.
Held cars are a fifth cause in a depot. Each one carries a reason code and a waiting-on value, which the out-of-service reason codes guide covers; this report only needs their hours.
The morning count and the car-hour rate: two numbers for two questions
An example, with made-up numbers to show the arithmetic. A 60-car roster has 1,440 car-hours in a day. Cars spend 780 hours in service and 180 staged, so 960 are available: 66.7 percent. They spend 228 hours charging and 60 in cleaning or inspection, so 288 are turning: 20.0 percent. And 72 hours are worked in a bay, 48 wait on parts, 48 sit on holds and 24 are away at a body shop, so 192 are down: 13.3 percent. The three add back to 1,440, which is the check.
Now the peak. Say the evening needs 44 cars on the road. The Federal Transit Administration defines the spare ratio as spare vehicles divided by the vehicles required for annual maximum service, so 100 required and 20 spare is 20 percent, and it says operators of 50 or more fixed-route vehicles should not exceed 20 percent.6 By that arithmetic this depot holds 16 spares on 44, or 36.4 percent. The guideline was written for federally funded bus operators, not for you. Our reading of the gap, an assumption: an electric roster carries its charging hours inside that margin, which is why the turning bucket deserves its own line.
What goes on the daily status report
- The roster count and its basis: every car, or assigned cars only, and any car moved to inactive since yesterday.
- Cars available at the morning release and at the peak hour, each beside the number required.
- Car-hours by bucket for the last 24 hours: available, turning, down, with the total.
- Down hours by cause: being worked, waiting on parts, held, away, unaccounted.
- The cars down longest, by number, with what each is waiting on and who owns it.
- Limited cars, if you count them as available.
The Defense Department updates its availability metrics, and each individual asset's not available time, monthly, and lets the services report in hours or days.1 A depot that releases cars every morning needs the same table daily, in hours. That cadence is our assumption. The count also belongs in the shift handover, so the incoming lead starts from the same numbers.
When an availability rate is the wrong tool
- You want to know whether rides were served. Availability says a car could have taken a rider, not that one was there. The Defense Department makes the same point: availability reporting is not to be confused with unit readiness reporting, and is only part of assessing it.1
- You run a handful of cars. Our assumption: below about ten, one car is ten points of the rate, and a list of cars with a status beside each tells you more than a percentage.
- The question is whether the driving system is fit to drive. That call belongs to the vehicle's maker and its trained staff; the report only records that a car was held.
- The chargers are what you are measuring. Port availability is its own number, covered in EV charger uptime.
- A high rate is being used as proof of a good day. A roster that is 95 percent available with half of it staged all day was oversized, not well run.
Questions AV fleet operations managers ask
What does availability mean for a vehicle?
How do you calculate an availability rate?
Is a car on a charger available?
Do spare cars and held cars stay in the count?
What is a good availability rate for an AV fleet?
What is the difference between waiting on maintenance and waiting on parts?
Sources
- U.S. Department of Defense: DoD Instruction 3110.05, Sustainment Health Metrics in Support of Materiel Availability (April 24, 2024)
- International Electrotechnical Commission: International Electrotechnical Vocabulary, IEV 192-01-23, availability
- U.S. Government Accountability Office: GAO-23-106217, Weapon System Sustainment: Aircraft Mission Capable Goals Were Generally Not Met
- Federal Transit Administration: National Transit Database (NTD) Glossary
- U.S. Government Accountability Office: GAO-22-104533, Air Force and Navy Aviation: Actions Needed to Address Persistent Sustainment Risks
- Federal Transit Administration: Circular C 5010.1F, Award Management Requirements (November 1, 2024)
Sources are public. XoomPark isn't affiliated with the companies referenced in this post.
Authorship
Written by XoomPark Editorial Team in collaboration with Claude Fable 5.2
Role breakdown: Topic selection and publishing by XoomPark Editorial Team; research, source checks against primary sources and draft prose by Claude Fable 5.2.