Submitted Sep 8, 2026
Ask most freight software what is happening with a shipment right now and it will tell you about tasks. A booking was confirmed. A customs entry was filed. A container was moved. Each of those is true, each is recorded, and none of them tells you whether anything is actually progressing.
That is the gap between digitising freight and running it. A task list is a record of activity. It is not a description of what a shipment is trying to do, and it cannot tell you whether it is closer to delivery than it was an hour ago.
The alternative is to make the unit of work the whole job rather than its pieces. Corvane calls that unit a mission.
What a Freight Mission Actually Is
A freight mission is the end-to-end workflow a shipment, driver or vehicle moves through to accomplish a single goal on a lane, expressed as a connected sequence of discrete actions that a system can orchestrate rather than merely record. Booking, pickup, customs clearance, linehaul, delivery and proof of delivery is one mission, not six unrelated tasks.
Daniel Okafor, SVP of Product and Engineering at Corvane Cargo, puts the distinction in one line: "A mission is a collection of tasks."
The reason that matters is not vocabulary. It changes what the software is able to do. "A task could be seen as a discrete action," Okafor says. "A booking, a customs filing, a load or an unload. Those are tasks. Missions bring them all together into an end-to-end workflow that can then be orchestrated."
A system that holds tasks can tell you a truck was booked. A system that holds missions knows the booking was step one of seven, knows what step two is, and can start it without waiting for a human to notice.
The question every mission answers
There is a useful test for whether something is a mission. Okafor points to the question every shipper is asked when booking with every forwarder in the world.
"What are you shipping is a classic first thing that every customer gets asked when they book freight, and effectively the answer to that is the mission, which then loads the series of nodes that then can become orchestrated by a network."
If your software cannot answer "what is this shipment for" in a single field, it is holding tasks.
Why Task-Based Freight Software Stalls
It documents the past instead of driving the present
The structural problem with most freight systems is not that they lack features. It is the direction they face.
"The majority of what is available today is documenting what has already happened," Okafor says. "A container clears a port, something happens, you have to remember to update the software to let the software know that the box went by. And so it's not helping your operation, you've just added another task for your team to do. Clear the container, and then go to the software and clear the container."
His conclusion is blunt: "It's just a digitisation process that adds work versus taking work away."
The tell is whether your team performs an action and then reports it, or whether performing it is the report. If someone is retyping reality into a system after the fact, the system is a ledger. Okafor's description of what that produces is unflattering and accurate: "Think about that as a really fancy spreadsheet. A really fancy database."
Nobody would accept this anywhere else in logistics
The rear-view model looks normal in freight only because it is universal there. Move it to any other high-throughput environment and it becomes absurd immediately.
"Can you imagine running a high-speed rail system rear-view mirror?" asks Marcus Hale, Corvane's Chief Marketing Officer. "Oh, train A31 came in from Osaka, right, okay, write it down." His point is not that it would be inefficient. "There's no way to do it. It just doesn't work efficiently, it just doesn't function."
Freight tolerates a standard that no comparably complex operation would accept, and the reason is historical rather than technical.
There is nobody left to do the data entry
The rear-view model has a dependency most shippers have not thought through: it requires a human to be standing there to record what happened.
"You move to the future where we need to be able to orchestrate movement autonomously, because we're taking humans out of the loop or interacting with autonomous vehicles and robots," Okafor says. "Then you can't be in the past. There's no one to enter the data for that truck that crossed the border, because there's no one there."
This is why the mission model is a prerequisite for automation rather than a nicer way to view it. Remove the data clerk from a task-based system and the system goes blind. Remove them from a mission-based one and the workflow continues, because the sequence was already defined and the sensors already feed it.
How Missions Are Built
Missions are assembled from nodes: discrete actions that map to things that really happen in freight. The architecture matters because it determines who can change it.
Signals come in. Live tracking sensors at ports, integrations with the WMS or TMS, driver input, telematics, and whatever sensing the lane has. The system learns a shipment has arrived without anyone telling it.
The mission is identified. The arrival is matched to a purpose: which shipment, which carrier, on time or late, full load or partial. This is the "what is this for" answer, resolved automatically.
Nodes are sequenced into a workflow. The mission loads the actions required for this specific purpose on this specific lane.
Decisions are made on the shipper's behalf. "AI is used to make decisions on behalf of humans," Okafor says, "qualified, accurate decisions, so that they can leverage their time more effectively, manage more of the operation, manage more of a network, by only focusing on the things that AI can't figure out." Humans handle exceptions, not the routine path.
The configuration belongs to the shipper. "It's configurable in a no-code way by the industry," Okafor says. "When you think about how every supply chain works different, we have that covered, and the greatest thing about that is you can over time configure these yourself."
That last point is where the model earns its keep. No two supply chains run identically, so a fixed set of workflows would fail everywhere except the lane it was designed for. Okafor is explicit that the interesting outcome is customers building sequences Corvane did not anticipate: "We're excited for our future when we see our customers creating missions which are unique combinations of nodes that we've never seen before."
Determinism, and the Part Nobody Can Program in Advance
There is an obvious objection. If a mission is a defined sequence, what happens when freight does something nobody defined?
This is where the mission model separates from simple workflow automation. A rail network is largely deterministic: a train leaves A and arrives at B, and the states in between are documented. Global freight is not. Hale raises the unknown-unknowns problem directly, the things you do not know that you do not know, and asks how a system accommodates them.
Okafor's answer is architectural rather than predictive. Because missions are built from a growing library of composable nodes rather than hard-coded end-to-end paths, the system does not have to anticipate every scenario. It has to be able to assemble a response from parts it already has.
"You're not building for every scenario," Hale summarises. "You're able to have a platform that's flexible enough to accommodate an unforeseen scenario."
Okafor adds the qualifier that matters: "That's fundamentally the AI revolution that we're seeing." With guardrails. Freight is a physical environment with safety and customs consequences, which rules out purely probabilistic decision-making. The system has to be flexible in what it can assemble and rigid about what it will permit.
A Diagnostic: Are You Running Tasks or Missions?
Can your system answer "what is this shipment for" as a single value, or does someone have to infer it from a sequence of records?
When a shipment clears customs, does the next step begin automatically, or does someone have to notice and start it?
If you removed the person who updates the system, how long before the system's picture of your freight is wrong?
Can a logistics manager change the order of steps for a workflow without raising a ticket with a vendor?
Does your software know that these six records are one job, or does it hold six unrelated rows?
When something unexpected happens, does the system have a path for it, or does it simply stop and wait for a human?
If the no answers cluster around the first three, you have a rear-view system. If they cluster around the last three, you have a workflow tool that is not configurable enough to survive contact with real freight.
Where This Sits in the Platform
Missions are the execution layer of a Corvane Control Tower, which is the distinction Corvane draws between a Control Tower and a traditional freight forwarding portal. A portal records shipment state. A Control Tower orchestrates cargo movement, and missions are the mechanism by which it does.
In practice that runs across air freight at arrival, linehaul and trucking for the moves in between, and live shipment tracking as the layer that makes the whole thing legible to a network manager rather than a single lane. The agentic AI layer sits on top of the mission backbone rather than beside it, which is why Okafor describes the AI as infrastructure rather than a feature.
Frequently Asked Questions
What is a freight mission? A freight mission is the end-to-end workflow a shipment, driver or vehicle moves through to accomplish a single goal on a lane, expressed as a connected sequence of discrete actions that a system can orchestrate. Booking, pickup, customs clearance, linehaul, delivery and proof of delivery is one mission rather than six separate tasks.
What is the difference between a freight mission and a freight task? A task is a single discrete action such as a booking or a customs filing. A mission is the collection of tasks that together accomplish the shipment's actual purpose, held as one connected workflow so that completing one step can trigger the next automatically.
Why does the distinction matter operationally? Because a system that holds tasks can only record that something happened, while a system that holds missions knows what should happen next and can initiate it. It is also the precondition for removing manual labour from freight, since a task-based system depends on a human being present to enter what occurred.
Can freight missions be configured without a developer? Yes. Missions are assembled from a library of nodes and are configurable without code, so a shipper can reorder or customise a workflow for a specific lane rather than requesting a change from the vendor.
How does a mission handle a situation nobody planned for? By being composed of reusable nodes rather than hard-coded end-to-end paths. The system assembles a response from components it already holds instead of requiring that every scenario be anticipated in advance, with guardrails constraining what it is permitted to do with physical cargo.
Where to Start
If your freight is digitised and your team is still doing data entry, the problem is the unit of work, not the software's feature list.
Corvane's Corvane Control Tower is built on the mission model described here. See how a Control Tower differs from a portal, or explore the platform and walk through air freight booking yourself with no sales call.
This article draws on Lights-Out Freight Episode 2 with Daniel Okafor, SVP Product and Engineering at Corvane Cargo.

