Railcar Fleet Management Software: Build vs Buy for Shippers

What the off the shelf railcar products actually do well

Start with what is not a product. Railinc is industry infrastructure rather than fleet management software. Umler is the equipment registry, the Car Repair Billing exchange moves repair invoices between parties, Car Location Messages carry movement events, and the Equipment Health Management System distributes wayside detector findings. All of that is essential and none of it tells you whether a specific car is costing you money this month. Treating Umler as a fleet system is a category error, and more shippers make it than would admit to it.

The genuine products are strong for the buyer they were designed around. RMI RailConnect and Bourque Data Systems were built primarily for railroad operations, and they do that well. A short line asks what is on my property and what can I bill. Lessor portals from the large leasing companies give you position, lease status and repair history for the cars they own, at no extra cost, and for a shipper leasing full service they answer most of the daily questions.

So the default answer is buy or use what you already have, and that is what we tell most callers. If you lease fifty cars full service from one lessor who handles maintenance, your portal plus a spreadsheet is proportionate and the recoverable money will not cover a project. The same holds at a hundred and fifty cars from two lessors, provided somebody actually opens the repair file each month.

The case for building starts when the questions change shape, and the questions change shape when you own the risk rather than rent it.

Where they stop: the repair line joined to the movement history

A fleet manager at a plastics producer controls nine hundred covered hoppers. On a Tuesday a repair billing file arrives covering thirty two cars. Each line carries a job code, a why made code, a responsibility assignment and an amount, drawn from the Association of American Railroads interchange rules. Some of those repairs are legitimately hers because they are wear items. Some belong to the handling railroad that caused the damage. Some duplicate work her own contract shop did six weeks earlier.

Separating them means cross referencing each repair against the car's movement history on the repair date to establish who held the car when the damage occurred. Nobody does that car by car. It would be a full time job with three screens and a set of industry manuals, and the window to dispute is not generous. So the bills get approved.

Meanwhile fourteen cars sit at a customer plant for eleven days with no storage recovery, six approach a tank car qualification date under the federal hazardous materials rules that will strand them if it lapses, and a lessor invoice still bills cars scrapped in March. Each number is modest. Together they are why the fleet budget is always over and nobody can explain it.

The reason no product does this is structural. Movement events arrive out of order, sometimes duplicated, sometimes contradicting each other when two railroads report the same interchange with different times. A system that takes the newest message and overwrites produces a car that appears to travel backwards. What you need is an event log with reconciliation rules per event type, a derived current position you can defend to a customer disputing dwell, and a join from each repair line to that history. Carrier products are not built to answer that, because it is not their question.

The arithmetic: cost to build against per car subscription pricing

Do this with your own invoice. Fleet management services in this space are usually priced per car per month, so take your annual figure and divide by cars. Then add the recoverable money you are currently leaving: disputed repair lines you never filed, storage and demurrage you never billed back under your tariff, mileage allowances you never reconciled, and lease invoices for cars you no longer hold.

On the build side, a first release at $70,000 to $150,000 spread over three years is $23,000 to $50,000 a year. Across one hundred cars that is $230 to $500 a car annually, which no subscription comes close to matching in the wrong direction. Across five hundred cars it is $46 to $100 a car. Across two thousand cars it is $12 to $25 a car, roughly a dollar or two a month, and the comparison stops being interesting.

The crossover sits near four hundred cars for a fleet on one commodity with one lessor, and falls to about two hundred and fifty once you own tank cars, because the qualification and inspection model is materially more complex and the cost of a stranded car is higher. Below one hundred and fifty cars, buy or use the lessor portal. There is no honest arithmetic that says otherwise.

One more figure belongs in the comparison. Fleets are sized for the cycle time you have, not the one you could have, so a reduction in customer dwell shows up as cars you do not lease next year.

What a custom build actually costs, migration and year two included

Fleet work sits in two price bands, drawn from what Digital Heroes has delivered across more than 2,000 projects. A first release covering movement event ingestion and reconciliation, current fleet position, dwell and cycle time reporting, and repair billing exception review runs $70,000 to $150,000 and ships in 12 to 18 weeks. A full platform adding qualification and inspection scheduling, lease and rider administration with invoice reconciliation, storage and demurrage recovery, shop scheduling and integration to plant or resource planning systems runs $180,000 to $450,000 phased across 6 to 12 months.

Data migration runs 10 to 25 percent of build cost, and in this category it is the lease riders and the maintenance history rather than the movement feed. Riders live as scanned documents with terms nobody has transcribed, and until they are structured data the invoice reconciliation cannot run. Budget that transcription as its own workstream, because it does not need a developer and it gates the highest return feature in the system.

Plan on 15 to 20 percent of build cost a year from the second year. That covers hosting, the movement feed format changes, new lessors and new rider structures, regulatory interval changes for tank cars, and the enhancements your fleet manager asks for once the exception queue is running.

What pushes you up the band: tank cars, multiple commodity types with different loading and cleaning requirements, the number of lessors, and integration to plant scheduling, which is where the largest operational value sits for a manufacturer and also where the most interfaces live.

The four situations where building wins

Regulatory fit. Tank car qualification and inspection intervals set out in the federal hazardous materials regulations are date driven and unforgiving, and a car that reaches its date is unusable until the work is done at a shop with capacity. Retrofit and specification deadlines for tank cars in certain services move on their own calendar. Holding those dates as first class data, projected against planned utilisation and grouped by geography so shop trips batch sensibly, is a compliance control rather than a report.

Scale economics. Above roughly four hundred cars the per car cost of a build falls below any subscription and stops growing as the fleet does.

A workflow that is your competitive advantage. If your fleet turns faster than your competitor's, you serve the same volume with fewer cars and less lease expense. That advantage is built out of cycle time by origin and destination pair, dwell split between customer plants and transit, and a named owner for every idle car. No vendor sells that because no vendor knows your customers.

Integration sprawl across three or more systems. Movement feeds from multiple railroads, the repair billing exchange, the equipment registry, lessor invoices and your own plant scheduling. The value is precisely in the joins between them.

Digital Heroes is the wrong firm for a fifty car full service lessee, and for any shipper who cannot name the person who will work the exception queue every week.

How to decide in a week, and the test that pays for itself

Monday, take last month's repair billing file and pick twenty lines at random. For each, find where the car was on the repair date and decide whether the responsibility assignment looks right. Time yourself. The hours that takes for twenty lines, multiplied by your monthly line count, is the reason nobody does it.

Tuesday, list every car that has been stationary at a customer for more than seven days and check whether anything was billed for it. Wednesday, take one lessor invoice and reconcile it line by line against cars you actually held and miles you actually ran. Thursday, pull your qualification due dates and count how many fall in the next six months without a shop slot booked.

Friday, add up what those four exercises found and compare it to the arithmetic above. If the total is under twenty thousand dollars a year and your fleet is under two hundred cars, buy, and put the effort into working the lessor portal properly. If it runs into six figures, the next step is a paid discovery phase rather than a build. Three to five weeks at a fixed fee, ending in a signed product requirements document covering the event reconciliation rules, the repair exception logic, the rider data model, permissions and acceptance criteria. You own that document and can take it to any other firm.

Digital Heroes starts with the signed product requirements document and writes no code before it. We contract through India LLP, United States LLC and United Kingdom LTD entities so the intellectual property assignment falls under your own law. The named team comes from more than fifty specialists and you meet them before signature. Check Clutch, Trustpilot, Fiverr Vetted Pro and D-U-N-S, then look at ShopScore, HeroCheckout and Section Vault, which we built and run.

Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.