Parts on the Shelf vs Truck on the Side: The Inventory Math That Saves 4 Hours Per Repair

The most expensive minute in the shop

A truck is in the bay. The mechanic has the work order. The diagnosis is clean. The repair is straightforward — replace the EGR cooler. The part isn't on the shelf. The next 4 hours go to: looking up the part number, calling three vendors, picking the one with the best price-and- lead-time, ordering it, finding out it doesn't ship until tomorrow, calling around for an alternative, finding it at NAPA 18 miles away, sending someone to pick it up, signing for it, getting back to the shop, opening the box, finding it's the wrong part, calling NAPA back...

Half the truck-days lost in the shop are not actual repair time. They are waiting on parts time. The part either wasn't on the shelf when needed, was the wrong part, or arrived later than promised.

Centrix's parts inventory layer attacks all three.

Stock what actually fails

The naive approach to parts inventory is "stock what we used last month." This works in steady-state but fails in transition — when fleet composition shifts, when a new truck make rolls in, when a vendor changes part numbers. Centrix's parts model uses three signals together:

• Historical parts usage from Shopmonkey work order data, by truck

make/model/year

• Predictive maintenance signals from the fleet's telematics +

service history (Samsara DTCs, oil analysis, mileage cohorts)

• Industry-wide failure data from the 70+ ML models maintained

across managed fleets — what fails on a 2022 Cummins X15 with 380K miles vs a 2019 Detroit DD15 with 520K?

These combine into a per-carrier stock recommendation: a recommended on-hand quantity per part with reorder point, lead time, and expected demand over the next 14 days.

The shop foreman sees: "You have 3 EGR coolers on shelf. Forecast demand 5 over next 30 days. Reorder 2 now (lead time 4 days from your preferred vendor), or you'll stock out around day 18." The decision is 30 seconds, not an hour of guessing.

Vendor lead time tracking

Most carriers' "vendor lead time" is a number on a spreadsheet from 2019. It's wrong, sometimes by days, sometimes by weeks. Centrix tracks actual lead time across every order: order placed, order shipped, order received. Over 90+ days, this builds an empirical lead time per vendor per part — accurate enough that the reorder point decisions are reliable.

The data also exposes which vendors are slipping. A vendor whose lead time has drifted from 3 days to 7 days over the last quarter is a relationship to renegotiate or replace, with data to back the conversation.

The wrong-part problem

Roughly 8-12% of parts orders result in a wrong part landing in the shop on the first try. Sometimes it's the supplier's fault. Sometimes it's the foreman's fault. Almost always it's because the vendor's catalog and the carrier's part numbering don't match perfectly.

Centrix's parts cross-reference layer maintains a normalized catalog that maps the carrier's part numbers to vendor SKUs across the top 12 parts vendors. The wrong-part rate drops to 2-3% within 60 days because the foreman is ordering against a normalized number, not a catalog page.

Per-truck parts cost tracking

A useful side benefit: every part that lands on a truck is costed back to that truck's lifetime maintenance ledger. This surfaces:

• Trucks costing 2-3x the fleet median in parts (candidates for sale

or trade)

• Truck make/model combinations with high parts cost (relevant at

next purchase decision)

• Drivers whose trucks consistently need more parts (sometimes a

driving-style issue)

• Parts categories that are silently growing (after-treatment parts

in particular have grown 40-60% across the industry over 5 years)

The data informs purchasing decisions, not just maintenance ones.

Bestpass and toll account reconciliation

A subtle but valuable cousin to parts inventory: toll-pass account management. Most carriers run Bestpass or PrePass for tolls. Most also have at least 2-4 trucks at any moment with a non-functioning pass — balance issues, account suspended, expired card on file, vehicle mis-registered. The trucks rack up cash tolls (1.5-2x the pass rate) plus violation fees because the carrier didn't know.

Centrix's toll account reconciliation cross-references Bestpass / PrePass account data with truck movement (Samsara) to flag:

• Trucks with passes that aren't reading
• Accounts in arrears (auto-renewal failed)
• Vehicles with the wrong axle classification on file (overpaying tolls)
• Trucks running through known toll zones without pass activity

Median first-year toll savings on managed fleets: $8K to $24K per 100 trucks purely from catching account issues that were quietly costing real money.

Pilot fuel card management

Same pattern, different system. Centrix's Pilot SFTP sync (per the fleet's existing Pilot relationship) reconciles fuel card transactions against truck movement and driver assignment. Fuel card fraud, account suspensions, and wrong-truck cards all surface as anomalies the fuel optimization role investigates.

(Note: fuel fraud detection is its own deeper article on the Accounting side — see Catching Fuel Fraud With ML. The Shop/Terminal view is more about card hygiene — the cards are working, the accounts are current, no truck is rolling without a working card.)

Mechanic time savings, where the dollars actually live

Putting numbers on it for a 100-truck fleet:

• Average 12 unplanned repairs per shop per month
• Average 4 hours of waiting-on-parts time per repair pre-Centrix
• Mechanic loaded cost ~$95/hr; shop overhead allocation ~$135/hr

effective

• Pre-Centrix waiting cost: 12 × 4 × $230 = $11,040/month per

shop in pure waste

Centrix's median outcome:

• Wait time per repair drops 60-75% (from 4 hr to 1-1.5 hr) via

shelf availability + better catalog + faster ordering

• Wrong-part rate drops 70%+ (from 10% to 2-3%)
• Net savings: $7,500-$8,500/month per shop

Plus the truck-day savings from each repair finishing faster — at $980/day fully loaded × 0.4 truck-days saved per repair × 12 repairs = ~$4,700/month of additional capacity recovered.

Combined: $12K-$13K/month per shop, or $144K-$156K/year on a single-shop 100-truck operation.

Where to start

If you're 30+ trucks running an internal shop:

• Connect Shopmonkey first if it isn't already. The work order data

is the foundation.

• Run a 30-day shadow on the parts model — Centrix recommends, the

foreman decides, the data improves.

• Turn on the vendor lead time tracking from day one. It calibrates

fast and produces useful renegotiation conversations within 60 days.

• Add toll and fuel card reconciliation in parallel — these are

independent of the parts model and pay for themselves quickly.

Book a parts and vendor review — bring 12 months of Shopmonkey work order data and your top 5 parts vendors. We'll show you exactly where the wait time and the wrong-part dollars are hiding today.