Centrix's Always-On Dispatch Engine: Routes, Fuel Stops, OD Permits, and 24/7 Monitoring Without a Human in the Loop

The four jobs your dispatch desk is doing badly

Look at any 8-hour dispatcher shift and you'll see the same four workflows competing for the same dispatcher's attention:

1. Picking the route — origin, destination, intermediate stops, hours-of-service constraints, customer preferences, weather 2. Picking the fuel stops — which Pilot, Loves, or TA on the route has the cheapest fuel, fits the truck's range, has the right amenities for the driver 3. Handling oversize/OD permits — which states the load crosses, which restrictions apply, which pilot cars are needed, which time-of-day windows are open 4. Watching the truck — is it moving? off-route? idling too long? running late?

Each one is important. None of them are creative work. They're well-bounded decisions a model can make reliably as long as it has the right data. The reason most carriers do all four manually is that the four data sources don't talk to each other — Samsara has position, the routing engine has the road network, the fuel card has prices, the state DMV has the permit rules.

Centrix's always-on dispatch engine sits across all four data sources and runs the workflows as one continuous loop. When everything is going to plan, the dispatcher sees nothing. When something needs a human decision, the dispatcher gets a notification with the context already attached.

Lever 1 — Route planning, HOS-aware

Every active load gets a route plan generated automatically when it's booked:

• Origin → destination via the network of roads the truck can legally use (no low

bridges, no weight-restricted roads, no hazmat-restricted routes for hazmat trucks)

• Customer-preferred lanes if specified (some customers require specific corridors

for security or service reasons)

• Driver's HOS clock factored in — the route is segmented into "drive shift" +

"10-hour break" + "drive shift" so arrival time is realistic, not theoretical

• Weather-adjusted — NOAA forecasts modify ETA when severe-weather windows

intersect the route

• Reload-aware — if a reload is booked, the route plans toward a stop position near

the next pickup, not wherever the BOL drops

The dispatcher sees the route. The driver sees the route in their Telegram bot in their language. The customer sees the ETA. All three views update in real time as conditions change.

Lever 2 — Fuel stop optimization

The single biggest controllable cost in trucking after labor is fuel. A typical Class 8 sleeper burns 12,000 gallons a year at $4.20 = $50,400 per truck per year. A 100-truck fleet is $5M of annual fuel spend.

The market price of diesel varies $0.40 to $0.80 per gallon across the network at any given moment. The cheapest station on a route is sometimes 2 miles off the planned path; sometimes 50 miles off. The math of "is it worth detouring for cheaper fuel" requires:

• The truck's current fuel level and range
• Distance to the cheapest station within range vs the next-cheapest
• Detour distance and time cost
• Driver-preference constraints (truck-stop chain, shower availability)
• HOS clock impact (does the detour push the truck past the 14-hour limit)

Centrix's fuel optimizer runs this calculation continuously. The driver sees the recommendation in their Telegram bot: "Refuel at TA Pilot Center, MM 217, $3.96/gal — 1.2 mi off route. You have ~180 gal range remaining." The dispatcher sees compliance with the recommendation in the morning brief.

Median per-gallon savings on managed fleets: $0.12 to $0.18/gal. On 12,000 gal per truck per year × 100 trucks = $144K to $216K of pure fuel savings per year without changing the fleet's mileage one bit.

Lever 3 — OD/oversize routing

Oversize and over-dimensional (OD) loads are higher-margin, but the routing complexity is brutal. A single OD move from Houston to Albany crosses 8 to 12 jurisdictions, each with its own permit office, fee schedule, lead time, and rule book. Some states allow weekend travel; others don't. Texas has different rules at different times of day. Pennsylvania has Sunday restrictions that vary by load class.

Centrix's OD routing layer ingests:

• Per-state permit rules for 50 jurisdictions, refreshed quarterly
• TruckerPath integration for low bridges, weight-restricted roads, and

time-of-day / day-of-week restrictions

• The truck's exact spec — height, width, length, weight per axle, axle count
• The load's exact spec — same dimensions plus tarp / escort requirements
• Pilot car availability in the carrier's vendor network

The dispatcher sees one decision: "this load is permittable in 6 hours, costs $1,840 in permits, requires 2 pilot cars in NJ, travels Mon–Sat between 7 AM and 6 PM. Net rate after permits and escorts: $4.18/mi." Versus bouncing between three websites for an hour to come up with the same answer.

Lever 4 — 24/7 monitoring + notifications

This is the lever that turns the always-on engine into a useful service. Centrix watches every truck, every minute, for things that need attention:

• Idle longer than threshold (3+ hours unscheduled) — could be a customer-side

delay or a driver-side problem

• Off-route by more than 5 miles — driver took a wrong turn, or there's a route

change worth knowing about

• ETA drift over 30 minutes — the customer needs to know
• Engine fault codes (Samsara DTCs) — Centrix triages these and only escalates

the ones that mean the truck is going to fail soon (the predictive maintenance model)

• HOS clock approaching limit — the dispatcher needs to plan a safe parking spot
• Fuel range under 100 miles — needs a refuel before the next planned stop
• Weather warning intersects the route — pre-emptive Twilio TTS call to the

driver in their language

When a watch fires, the dispatcher gets a notification with full context: which truck, which load, what happened, what to do. Notifications come via Telegram, SMS, or email per dispatcher preference.

The notifications are calibrated to be useful, not noisy. A good calibration target is 3 to 6 actionable notifications per dispatcher per shift — enough that the dispatcher trusts them, not so many they get tuned out. Calibration improves as the system learns which notifications you act on vs which you dismiss.

What the math looks like

A 100-truck fleet running this engine end-to-end:

• Fuel optimization: $144K–$216K/yr saved at $0.12–$0.18/gal across 1.2M gallons
• Idle reduction caught by monitoring: 22% → 14% idle ratio = $1,400/truck × 100 =

$140K/yr saved in fuel alone

• Off-route catches and ETA recoveries: ~$50K/yr in customer trust + accessorial

recovery

• Monitoring labor consolidation: 1–2 dispatcher seats freed = $80K–$160K/yr
• OD permit time savings: 90 min per load × 4 OD loads/wk × 50 wk = ~$45K/yr

Total impact per 100 trucks: $460K–$610K per year.

Where to start

If you're a 30+ truck fleet running dispatch on manual routing and reactive monitoring:

• Connect Samsara + your fuel-card provider first. The engine needs position + fuel

data to be useful.

• Run fuel optimization in shadow mode for 30 days — Centrix recommends, the driver

decides. Compare adherence and dollar impact after 30 days.

• Turn on monitoring + notifications after 60 days, once your dispatcher has built

trust with the system. Start with high-value alerts (HOS approaching limit, ETA drift over 30 min) and expand from there.

Book a demo — bring 90 days of Samsara position data and your top 5 lanes. We'll model the route + fuel + monitoring savings on your actual fleet.