A.R. PAGET TECHNOLOGIES
Sign inRequest a trial
Network Optimizer

The optimal plan, certifiably proven.

Which lanes to run, at what frequency, and what each leg costs — solved to proven optimality at up to 500 terminals. Four modes are modeled — road, rail, air and ocean — and where road and rail meet, the optimizer picks the mode lane by lane.

See the workspace Request Demo
Run record
2026-08-11
{{ r.k }} {{ r.v }}
The plan ships with this record attached. A proven label may not sit beside a non-zero gap; that is checked automatically before anything ships.
{{ p.n }}{{ p.unit }}
{{ p.label }}
{{ p.basis }}
Cost benchmark · optimized plan vs two baseline plans of the same freight
500-terminal benchmark · road + rail · verified solves
45% lower cost than routing each shipment on its own — $4.65M less every week on the modeled network

One week of the same freight, priced identically: the optimized plan costs $5.71M against $10.36M when every shipment takes its individually cheapest path — solved end-to-end through the shipped engine and proven optimal, with the same freight served in both plans. The percentages are the transferable statistic; the dollar magnitudes scale with the modeled network’s volume, so read them as benchmark economics, not a measured customer result.

59% lower cost than keeping every route running

Against a plan that keeps the full network open and paid for whether each route is needed that week or not, the savings run larger at every size — 49% to 59%. The shipment-by-shipment comparison is the harder one, so it stays the headline.

One week of the same freight, priced identically the only difference between the columns is how the network is planned
The proof story what each solve searched, and what the engine could prove
Revenue you would need to win instead saving ÷ operating margin = revenue so $242.0M at a 3% operating margin needs $8.07B of new business
Network Keep every route running Route each shipment on its own Optimized plan Savings
{{ w.size }} {{ w.all }} {{ w.own }} {{ w.opt }} {{ w.save }}
Keep every route running
The full network stays open and paid for whether each route is needed that week or not.
Route each shipment on its own
Every shipment takes its individually cheapest path; nothing is consolidated.
Optimized plan
Network Optimizer chooses which routes to run and how freight shares them.
Savings
Quoted against the harder comparison — shipment-by-shipment routing. Savings against keeping every route running are larger, 49–59%.
Network Candidate lanes Route options evaluated Result Solve time
{{ r.size }} {{ r.lanes }} {{ r.opts }} {{ r.res }} {{ r.t }}
“Proven optimal” appears only where the run’s receipt shows an achieved gap of 0.0. Every figure in this section traces to the run receipts to the cent; the page shows rounded millions.
Network Annual freight spend, optimized Annual savings At 3% operating margin At 5% operating margin At 8% operating margin
{{ p.size }} {{ p.spend }} {{ p.save }} {{ p.m3 }} {{ p.m5 }} {{ p.m8 }}
The savings
Annualized at 52 steady-state weeks of the same freight, quoted against the harder comparison — routing each shipment on its own. Against keeping every route running the annual savings are larger: 49–59%, $91.5M to $434.1M.
The magnitudes
A synthetic benchmark network solved end-to-end through the shipped engine. The percentages are the transferable statistic; the dollar magnitudes scale with the modeled network’s volume — benchmark economics, not a measured customer result.
The margins
Operating income over revenue, and yours to pick — ATRI 2025 puts truckload under 1%, tank 4%, LTL 11.6%. These use company-average margin; freight added to an existing network usually earns better than average, so the revenue needed is likely less.
The timing
Purchased transport lands at once; owned capacity as it is redeployed.
What the savings are actually worth

A dollar saved on line-haul is a dollar of operating income.

Line-haul sits inside your operating ratio — operating expense over revenue, the number carriers report every quarter. A dollar you stop spending there becomes a dollar of operating income, without a new customer, a new lane or a new truck. So the comparison worth making is not your freight spend — it is growth.

Network console

Set a delivery promise and watch the plan re-solve.

Each line is a real freight lane — intermodal, air or road — drawn at the volume that moves on it. Drag the slider to tighten what you promised the shipper. Road transit is bound by hours of service, so a lane past its rail cutoff drops back to road in amber, and a lane no legal road run can reach at all moves to air in blue. Down to 96 hours the panel re-reads the solve; below that it reports derived air transit instead.

{{ c.tag }}
{{ c.body }}
The Network Optimizer workspace

Open any lane and see how its routing was determined, dollar by dollar.

Network Optimizer plans road, rail, air and ocean freight in one model. The current benchmark solves a 500-terminal network to a proven optimum — a zero gap, meaning no cheaper plan exists for that freight. Pick a lane to see the two costs the optimizer weighed against each other, built up per container, and what the choice is worth across a week of volume.

Account overview Sample carrier · 102 terminals · week of Aug 10, 2026 · select a metric to see how it was derived
{{ drill.title }} Derivation

{{ drill.body }}

{{ drill.formula }}
{{ r.k }} {{ r.v }}

This account's baseline is a run-everything plan that operates every candidate lane directly, which is why its rate sits far above the figures earlier on this page — those measure a solved network against a normal plan on the same lanes. Different question, different baseline.

Rate card
Cost model
cost = base + per-mile (by equipment) × miles
Truck feeder
{{ t.k }}
{{ t.v }}
Air spine
{{ t.k }}
{{ t.v }}
A movement carries one cost. Service class does not change it — timing propagates; cost does not.
Why the card matters

This is the whole price list, and every figure in the plan is built from it. There is no blended average and no cost per unit hiding a mix of lanes — a movement's charge is a base plus a per-mile rate for its equipment, times the miles it actually runs. That is what makes a saving checkable: change one number here and the plan re-prices in front of you.

It is also what keeps the optimizer honest. Air costs roughly sixteen times a dry van per mile, so the engine only flies a lane when the date leaves it no ground option — never because air scored better on a blended rate.

Rate matrix — $ per mile by equipment
Equipment Mode $ / mile {{ m.eq }} {{ m.mode }} {{ m.rate }}
Ocean and rail are bought capacity, priced per container by the partner settlement rather than by this carrier's base + per-mile card, so they carry no per-mile figure. Both levels are provisional — placeholders pending a sourced rate, never tuned to make a plan look cheaper.
Worked example — truck · BOS → NYC
53′ reefer · Standard · 223 mi
$95.00 base + $2.65/mi × 223 mi = $685.95 per load
Worked example — air · CHI → ATL
Air ULD · 689 mi flown
$6,000.00 air base + $33.00/mi × 689 mi = $28,737.00 per load
Air differs from truck by rating per aircraft-mile and carrying no driver-hours block — not by any service premium.

How we measure

Every number on this page traces to a run

Each one carries the network, the baseline it was measured against, and the gap certificate — and we will re-run any of them in front of you on your own data.

{{ e.title }}

{{ e.body }}

{{ e.receipt }}
Our products
Network Optimizer You are here
Design, schedule and optimize your network

Which lanes to run, at what frequency, on what cadence, at what cost — solved to a proven gap and kept optimal against the service commitments you already sold.

Long-haul, intermodal and middle-mile · road and rail · multi-tenant
Veridiance
Your lanes are optimized but your paperwork isn't.

Veridiance automates the entire document management process for freight forwarders and brokers.

See broker automation

Bring us a network. We'll optimize it and show you the receipts to prove it.

A lane file and minimal paperwork is all we need to prove the value of Network Optimizer. Even with the bare minimum, we will optimize your lanes against your current plan as the stated baseline — and show the receipts behind every number.

Request a trial
Your request is ready to send

Your mail app should have opened with this filled in, addressed to hello@arpaget.com. If it did not, copy the details below and send them from anywhere.

{{ sentText }}
Copy the details Done
Request a trial
Thirty minutes, on your own lanes. We will come prepared, so the more you tell us the less of the call is discovery.
About you
About your operation
What you are looking for
{{ demoHint }}
Cancel Send request