Under the hood

An algorithm that tells you the truth

Most load planners promise "optimal". We promise something better: the same plan every time, hard rules it never breaks, honest failure when cargo doesn't fit, and published numbers for how well it packs. Here is exactly what it does — and what it deliberately leaves to your judgment.

In the engine

What the algorithm does

Deterministic

Same manifest, same vehicle, same result — every single time. No randomness, so a plan you approved yesterday is the plan you get today.

Honest about failure

Cargo that doesn't fit is counted and shown, never hidden. "46 of 50 placed" is a result you can act on; a silent 50/50 you can't trust is not.

Unload order first

Later stops load deep, earlier stops stay reachable, and stacking never buries a pallet you need at the first drop.

Real stacking rules

Layer caps, load-bearing weight limits, support area, vehicle height — the same physics whether the engine places a unit or you drag it by hand.

Axle & CoG steering

With your vehicle's axle data, the engine packs and repairs the layout so the cargo's centre of gravity lands inside the safe range — lengthwise and sideways — and flags what it can't fix.

Checked for every leg

On multi-stop routes it plans and re-checks axle loads after every drop, because a plan that's legal at departure can go wrong at stop two.

By design

What it deliberately does not do

It doesn't claim "optimal"

True optimality for 3D loading is not computable at interactive speed — anyone claiming it is selling adjectives. We publish benchmark numbers instead.

It doesn't weigh your cargo

Every number comes from your inputs. Wrong weights in, wrong plan out — which is why verification stays with your team.

It doesn't replace your responsible person

Axle loads and load securing must be checked by a qualified person before departure. The report makes that check fast; it does not make it.

It doesn't tip or flip cargo

Rotation is flat (90°) only — correct for pallets, deliberate for stability. "This way up" stays up.

It doesn't certify legal compliance

Limits are the ones you configure for your vehicle. We check your plan against your numbers — we don't interpret transport law.

Evidence

How we measure it

Every change to the engine must pass two published benchmarks before it ships — a geometric suite scored by an independent auditor against a frozen baseline, and an axle-compliance suite scored by the same physics the app shows you.

Geometric fill (synthetic suite)

26 of 35 known-feasible cases pack completely — a maximal-rectangles reference engine manages 25 of 35.

Axle & CoG compliance suite

10 of 10 scenarios land within limits across every route leg — up from 4 of 8 before CoG steering shipped.

Measured July 2026. These numbers change only when a benchmark run changes them.
Your part

Plan here. Verify before the wheels turn.

Snugload is a planning tool, not a scale and not a legal authority. It computes estimates from the dimensions, weights and vehicle data you enter — if those are off, the plan is off. Use it the way professional dispatch teams do: plan here, then have your responsible person verify actual weights, axle loads and load securing before the vehicle moves. The report is built to make that check fast, not to replace it.

Snugload gives planning estimates from the data you enter. Final weights, axle loads and load securing must be verified by your responsible person before departure — the way every good dispatch runs.

Your next manifest is the best place to start.

Import it, hit Auto-position, and see what fits — and what doesn't. No card, right in the browser.

← Back to Snugload