Planning a multi-drop trailer load

Six consignments, four consignees, one trailer. The freight fits — that part was never in doubt, and a solver settles it in a second. What decides whether the run works is whether the pallets for stop two are reachable when the doors open at stop two, and whether the vehicle is still a sensible vehicle after they have gone.

That is a different problem from packing, and it is decided before anything is loaded. This page walks the decisions in the order you hit them, and starts with what getting the order wrong actually costs — because that tends to be told as one story when there are two, and which of the two you are facing decides how much of the plan you should spend on it.

What actually happens at the second stop

The standard version of this story is that a badly sequenced trailer means a wasted round trip to the depot. That version is real, and it is worth taking seriously. When the consignment for stop two is genuinely unreachable — nothing can be shuffled because there is no free floor left, the goods cannot safely be restacked at the kerb, or the receiving site simply has no forklift that can get at it — the delivery does not happen. Then it is a second visit, a second slot, a missed window and a conversation with the customer, and that cost dwarfs whatever the sequencing was traded against.

What it is not is the only version, and treating it as the only one gets the calculation wrong in the other direction.

I worked internal logistics for a global manufacturing company, receiving and unloading a lot of trucks, and this problem came up constantly. What I watched drivers do, when the piece I was expecting was behind two other consignments, was reorganise the load themselves on the spot with a pallet jack — and it is nothing short of an artform. Quickly grabbing two massive cargo blocks, moving them to the back of the trailer in order to get at something and squeeze it through the door with only a centimetre of margin on each side, and making it look like natural instinct and perfected technique. Before I had even managed to position my forks to grab the cargo I was receiving, they had restored order in what was left.

So sometimes the time cost is far smaller than it sounds, and sometimes it is the whole delivery. Which of the two you get depends on the cargo, the trailer, the equipment on the dock and who happens to be driving — and a planner in an office controls none of them. That is the honest case for planning the order: not that every mis-sequenced trailer fails, but that the downside is variable and its bad end is expensive, so it is not something to leave to chance and goodwill.

The costs that are easiest to miss, though, are not the driver’s time at all. There are two of them.

The first is the receiving site’s capacity. At the site where I worked there was a queue of trucks waiting to be unloaded, one spot to park in, and one forklift. A load sequence that looks fine on paper causes a queue of vehicles the moment the consignee has to become involved in rearranging the truck. That is the constraint the planner in the office cannot see, and it is the one that actually bites: not the driver’s twenty minutes, but the twenty minutes multiplied across everyone behind him.

The second is liability, and it is the one that changed how I think about this. My inclination was always to help — get the driver unloaded and reloaded on the kerb as fast as possible, get the queue moving, be rid of the problem. Until one day a senior colleague saw what I was about to do and pulled me aside to explain it. We could not take responsibility for other people’s cargo being unloaded and reloaded, or for how it was restacked afterwards. If anything happened, we would be in a very uncomfortable position. And we would have locked up the line of trucks behind us while doing it.

So, no more Mr Nice Guy after that day. We required, always, that drivers make their own cargo accessible for unloading from their vehicle. If there was a queue behind them, we asked them to reload themselves — but not at our only unloading station — and rejoin the back of the queue.

What happens to your trailer at stop two therefore depends entirely on who you meet at the receiving dock, and you do not control that. So what you are avoiding is wider than the round trip. It is also the safety of people carrying an unintended workload of loading and unloading cargo they have no business touching in the first place. Reloading is more manual handling, and more manual handling raises the risk of damage and injury, however skilfully it is done.

That is a strong case for planning the order deliberately: a failed leg at the top end, queue time and exposure at the everyday end. It is not a case for treating discharge order as the only thing the plan is for. Load order is one constraint among several, it is worth real effort, and how hard you enforce it on a given job is a judgement — which is what most of the rest of this page is about.

The stop order is your decision, not the plan’s

Worth being blunt about this before any settings, because “multi-drop planning” reads as route optimisation to most people arriving from a search, and that is not what this is.

Cargo-Planner sequences a load to a delivery order. It does not work one out for you, and that is true of every mechanism on this page. Routing is the clearest case: it reads the destination you have written on each item and orders the plan so the trailer discharges in the reverse of the order it was loaded — last on, first off. It matches destinations as text, not as geography: any string works, because the sequencing runs on string equality rather than on distance between points. The location picker behind the Destination column is map-backed for convenience of entry; nothing about the sequencing reads coordinates.

Which means route feasibility, customer time windows, driving hours, and whether the first drop can be made from the driver’s side of the road are all yours. You decide the stop order. The plan enforces it — as strictly or as loosely as you tell it to.

For a forwarder that is usually the right division of labour, because most of what fixes the order never reaches the software anyway — the consignee who only accepts deliveries before ten, the site with no forklift, the drop that has to be first because it is the only one with a booked slot.

Getting the drops into the load list

The consignee data arrives on the customer’s spreadsheet, in the customer’s column names. Importing via Excel takes it as it comes: paste or upload the file and use the column mappings to say which column is which, so nobody re-types six consignments to change a header.

The column that matters here is Destination. It is free text and accepts a country, a city, a postcode or a street address — or a stop code you invented, which is often cleaner.

Because it is free text, the convention is yours to enforce, and there is only one rule that matters: one string per stop, spelled identically on every item going to that stop. “Malmö” and “Malmo” are two stops as far as the sequencing is concerned. If the customer’s spreadsheet has the consignee name in one column and the town in another, decide which one is the stop and map that, rather than hoping they agree.

Priority is a mappable column too, and on a short run it is often the one to map instead, for the reasons in the next section but one. Nothing stops you filling in both; only one of them will decide the order.

If you do sequence on destination, routing also needs a Port of loading on the load list itself, because that is where the route starts. It sits with the Port of destination in the Edit loadlist info panel on the loadlist data page, reachable from the three-dot menu in the toolbar.

Destination and consignment ID do different jobs

This is the fork that catches people, because both fields look like they are about “which shipment is this”.

Destination controls the order. It is what routing sequences on.

Group / Consignment controls togetherness. Items sharing an ID are loaded together, and by default items with different IDs are loaded into different containers.

On a four-drop trailer you usually want both, and the default on the second one is the one to watch: six consignments with six IDs, left alone, means the plan reaches for more than one trailer. If they are meant to share a vehicle, that default has to change. The Shipments / Groups settings are where it lives, along with the other three decisions of the same family — whether shipments may be mixed in a container at all, whether a shipment must sit contiguously without other cargo in between, and whether a shipment travels whole or not at all.

That last one is worth a moment for a forwarder. A consignment that arrives 80% complete is often worse than one that did not travel, because the consignee cannot do anything with it and you now have a split delivery to administer. If that is true of a customer, say so in the settings rather than in an email.

Pick the mechanism that matches the constraint

“Multi-drop” makes routing sound like the obvious setting, and it is not always the one you want. There are four ways to decide load order here, and they are alternatives rather than layers. Take the most direct one that expresses the constraint you actually have.

Priority on each item is the plainest of them, and the docs say as much: priority is often the cleanest choice when the load sequence is a business decision rather than a route. Higher priority loads before lower, and both load before items with no priority set. It is an ordinary cargo property — it arrives on the same spreadsheet as the weights and dimensions, it needs no port of loading, and a planner reading the load list can see the intended order without having to know which settings are switched on. For three or four drops whose order you have already fixed for reasons that have nothing to do with geography — the booked slot, the consignee who only receives before ten, the site with no forklift — writing 4, 3, 2, 1 down a column is a shorter path to the same trailer than modelling stops.

Strict priorities is the hard version of the same field: a lower-priority cargo is never loaded before a higher-priority one, even if the higher-priority cargo cannot be loaded at all.

Routing is the right tool when the stops are genuine delivery destinations and there are enough of them that keeping priority numbers correct becomes bookkeeping — when consignments are added, cancelled or moved between vehicles, the destination on each item stays true while a hand-numbered sequence has to be renumbered. The toggle is Enable Routing, in Load settings → Primary settings. It is off by default and only appears on road-type load plans; on sea and air lists the Destination column still exists for the map view and reporting, but there is no routing toggle.

What routing takes over is worth knowing precisely, because it is not everything:

• When routing is on, it overrides Priority and the Strict priorities setting for the purposes of load order. You cannot hand-tune the sequence with priorities underneath it and expect both to apply.
• Priority still influences which items ship if not everything fits. That has not been taken over, and it is a different question from what order they go in.

So routing and priority are a choice, not a stack. Decide which of them owns the sequence on this plan, and if the answer is “mostly the customer’s instructions, with geography as a tiebreak”, that is an argument for priority.

Must be loaded before is a pairwise rule between two item selectors. Use it when the requirement is “this goes in before that” and nothing else about the load needs sequencing — one machine that has to go against the headboard, one consignment that must not end up behind another.

Order by sorts the loading order by any cargo property — weight, SKU, a custom column. Reach for it when the order follows something the cargo already knows about itself.

The sharp edge: those two custom rules are independent of the routing toggle, and the plan does not merge them with destination sequencing. Turning routing on and also writing a Must-be-loaded-before rule is not finer control over the same sequence; it is two mechanisms that do not combine. Pick one.

How hard the order should be is a separate decision

Choosing the mechanism settles what the order is. It does not settle how much the plan is allowed to give up to honour it, and that second decision is where most of the judgement lives. Cargo-Planner will hold a sequence rigidly or treat it as a preference, and the difference in utilisation between those two answers can be a whole trailer.

The rigid end is the four mechanisms above, plus the Shipments / Groups settings already described — no mixing in a container, contiguous shipments, whole shipments only. Those are on/off. They do exactly what they say and they cost whatever they cost.

The loose end is where the interesting options are, and they are custom rules:

• Group by, with Destination as the property, keeps each stop’s cargo together — and its reachability options are the part worth knowing about, because they model the person doing the unloading rather than the geometry of the boxes. Reach length is how far a loader can reach across another group to get at cargo, which is further with a forklift than by hand. Reach to back of cargo decides whether they have to get at the far side of an item to move it. Access width is how wide a corridor has to be before a loader can drive into it rather than reach over. Set together, they answer the question the first section of this page is really about: can stop two be got at without anybody touching stop three’s freight? Strict loading then decides whether a group must be finished before the next one starts — which reduces how far groups spread across containers, and, as the docs put it plainly, uses more space while doing so.
• Minimize mixing, again on Destination, is the same idea at container level and with the trade-off made explicit. Items with different destinations should not share a trailer — unless mixing them buys enough to justify it. The threshold conditions are what make it usable: allow mixing only if it brings the container count under a limit, only if the total cost exceeds one, or allow it freely but apply a cost penalty to mixed containers so the solver prefers clean ones. That is the closest thing here to writing down “a clean stop separation is worth something to me, but not a second vehicle” and letting the calculation settle it.

Neither of those is a softer version of routing — they control what shares a trailer and what stays reachable, not the discharge sequence itself, and on a real multi-drop plan you will often want one of each. The point is that “sequenced” is not a switch. It is a dial, and setting it is the planner’s call, not the solver’s.

Last-stop-first is a rule, not a law of physics

Every article on this subject repeats “load last-stop-first”, and it deserves its reputation: it is the right default, and a trailer built to it will usually discharge without incident. What it is not is the only constraint, and on real freight it rarely gets to be.

The cargo may have weight-distribution requirements. It may need to be secured in a particular way. Some goods cannot safely be moved around during unloading at all. The first delivery may have no forklift while a later one has a full warehouse team — which argues for making the hard stop easy rather than the first stop easy. Certain cargo may have to stay segregated from other cargo. Or the vehicle itself may have loading constraints that make a theoretically perfect discharge sequence physically impossible.

What beats last-stop-first is a plan built on the actual operational constraints of the whole route. Sometimes that means giving up some utilisation to get a clean unloading sequence. Sometimes it means dividing the trailer into practical zones. And sometimes — particularly where the alternative is repeated unloading and reloading, extra labour, delays and missed delivery windows — a second vehicle is genuinely cheaper than one badly sequenced vehicle.

I am not going to give you a percentage for how much utilisation a clean sequence is worth, because the honest answer is that it depends on what the alternative costs at the receiving end, and that varies by consignee. It is a judgement, made per job, and the thing to avoid is making it by default in either direction. Treating utilisation as the only objective is the obvious mistake; treating discharge order as the only objective is the same mistake in different clothes, and it shows up as trailers ordered that did not need to be. A slightly less full vehicle that can be loaded, delivered and unloaded predictably is often the better plan, and the space that buys the sequence is not wasted space — but it is space, it is on the invoice, and it should be spent knowingly.

The route is a sequence of load states

Here is the part that a static loading calculation does not naturally capture, and I want to be careful about how strongly I put it.

Cargo-Planner reports axle weights for the plan as built, once cargo is placed — gross and net, per axle set — against the limits you set per axle set on the equipment. That answers “can I load this trailer legally”, which is the question that gets asked.

The harder question is whether it stays acceptable for the rest of the route. Once you start removing heavy consignments, the centre of mass moves. A trailer that begins with a sensible distribution can end up carrying most of its remaining weight over one part of the chassis. The fully loaded trailer is only the first state of the journey; for a multi-drop vehicle, the route is a sequence of different load states, and only the first of them is the one you calculated.

I have not personally watched a trailer that was legal at departure go out of axle tolerance on an intermediate leg, and I would be careful not to overstate how systematically this is managed in everyday operations. Do I think every forwarder recalculates axle loads after every delivery? No. For ordinary, relatively light freight that would often be disproportionate.

But for heavy cargo, concentrated loads, specialist equipment, or a route where the drop sequence is fixed and cannot be adjusted afterwards, it is absolutely worth planning around. Nothing recalculates it for you between stops — if you want to see a later state, you plan it as its own load with the remaining consignments and read the axle weights off that. Doing it for the one drop that removes most of the weight is usually enough to know whether the question matters on this job.

What the driver gets at each stop

The plan has to leave the office in a form the yard can build and the driver can work from, which for a multi-drop load means the loading sequence rather than a picture of the finished trailer.

Build a load plan layout with the loading instructions block, which gives step-by-step instructions by layers, by stacks or by a custom property — on a routed load that sequence is the reverse of the discharge order, which is the whole point of having set it. Export it as PDF or Excel for the dock, and share it as a read-only link with the consignee or the carrier so the receiving site knows what is arriving and where in the trailer it sits.

That last one is worth doing precisely because of the first section of this page. A receiving site that knows the freight for them is at the back does not have to discover it with a queue forming. If none of this is familiar yet, creating a load plan walks the four steps, and why loading instructions matter covers what makes them survive contact with the loading bay.

When it stops being a multi-drop problem

Three ways this job turns into a different job:

• One of the drops is oversized. Once a consignment leaves the legal envelope, permits and axle limits drive the plan and the sequence becomes downstream of them — see moving out-of-gauge cargo by road.
• The question is a count, not an order. If what you actually need is how much vehicle this freight needs, start with how many pallets fit in a truck or loading metres.
• You are not sure a load plan is the right artefact at all. What a load plan is covers the concept and the audiences it has to serve.