Which freight lanes should an AI logistics agent start on?
Key Takeaways
- Wilson, by Cartage, is an AI logistics agent. Most customers start Wilson on a few lanes and expand from there.
- Cartage doesn't run pilots. The first lanes are the start of live freight work, not a separate trial program.
- Good starting lanes repeat often, use approved carriers, and arrive with complete order data.
- A starting lane should still produce some exceptions. A lane with no surprises proves very little.
- Lanes with open rules, such as unclear accessorial policy or a disputed carrier, should wait.
- Cartage helps identify the highest-impact lanes during discovery.
- Each lane should have a named owner who handles decisions above the approval limits.
What makes a freight lane a good place to start?
A good starting lane is busy, repeatable, and already governed by clear rules. It moves often enough to show results. Its carriers are approved. Its orders arrive with the data a shipment needs. And one person owns the decisions that fall outside the rules.
A lane is an origin-to-destination movement, such as a plant in Ohio to a distribution center in Georgia. But the same lane can carry very different work. One LTL order ships on a standard pickup. Another needs a liftgate and a delivery appointment. A third goes FTL at quarter end. So a lane is really a pattern of work, not just two points on a map.
Here is a typical choice. A building products manufacturer ships about 25 loads a week across 40 lanes. Six lanes carry more than half the volume. Those six use four approved carriers and the same order format. A seventh lane is the team's biggest headache. It has a new customer, unclear delivery rules, and a carrier the team is still negotiating with. The six steady lanes are the right start. The seventh comes later, once its rules are settled.
The most painful lane is tempting. It feels like the place where help matters most. But pain often comes from missing rules, not from missing hands. An AI logistics agent follows rules well. It cannot invent rules the company has not made yet.
Five screens for choosing the first lanes
Run every candidate lane through five screens before it goes live. A lane that passes all five is ready. A lane that fails one waits until that gap is closed.
Volume. Does the lane move often enough to show a pattern? A lane that ships twice a quarter gives the team little to review.
Carriers. Are the carriers on the lane already approved? Wilson books only carriers the company has approved, so a lane that depends on a new carrier waits for that approval.
Order data. Do orders arrive with weight, dimensions, pickup window, and delivery details? Missing fields turn every load into a question.
Rules. Are the rate targets, accessorial rules, and approval limits written down? If two coordinators would handle the same load differently, the rule is not ready.
Owner. Who decides when a load falls outside the rules? Every lane needs a name, plus a backup.
The order matters. Volume comes first because a lane with too few loads cannot show whether the work changed. The owner comes last, but it is the screen teams skip most often.
Lane types compared: start now, start soon, or start later
Most lanes fall into one of four types, and each type has a clear starting decision. The table shows how to sort them.
| Lane type | What it looks like | Starting decision | Why |
|---|---|---|---|
| Steady core lane | High volume, approved carriers, standard orders | Start now | Shows results quickly and builds trust in the rules |
| Busy lane with known exceptions | High volume, regular late pickups or appointment changes | Start now | Tests exception handling on real loads |
| Mixed-mode lane | Same origin, both LTL and FTL depending on order size | Start soon | Ready once the mode rule is written down |
| Unsettled lane | New customer, unclear rules, carrier not yet approved | Start later | Needs business decisions before any software can help |
Two of the four types belong in the first group. That mix matters. A steady lane alone shows that routine booking works. A busy lane with regular exceptions shows what happens when a pickup slips.
When to hold a lane back
Hold a lane back when its open questions are business decisions, not software questions. Three signals usually mean a lane should wait.
- The carrier list is in flux. The team is still choosing between carriers, or one carrier is under review. Wait until the approved list is settled.
- The customer rules are unwritten. Delivery appointments, unloading terms, or notification preferences live in one person's head. Write them down first.
- The rate target is disputed. Finance and logistics disagree on what the lane should cost. Wilson negotiates rates against the company's rate targets. It needs one agreed target.
Holding a lane back is not a failure. It is a list of decisions the team needs to make anyway. Once those decisions are made, the lane joins the next group.
A sample lane plan for the first weeks
A first lane plan names each lane, the reason it is included, and what it will show. Here is an example for a distributor with 30 active lanes.
| Lane | Mode | Reason to include | What the team will review |
|---|---|---|---|
| Dallas DC to Houston customers | LTL | Highest weekly volume, three approved carriers | Quote speed and booking confirmations |
| Dallas DC to Oklahoma City | FTL | Regular volume, one primary carrier and one backup | Tender acceptance and backup carrier handling |
| Supplier in Monterrey to Dallas DC | Cross-border | Frequent late pickups | Exception response and team notifications |
| New retail customer in Phoenix | LTL | Not included yet | Waiting on delivery appointment rules |
The plan is short on purpose. Three lanes cover three modes and one known exception pattern. The fourth lane is written down so nobody forgets why it waits. After the team has reviewed the results, it can decide what to add. For that decision, see After 30 days on a few lanes, should the AI logistics agent expand?.
How Wilson, by Cartage, starts on a company's first lanes
Wilson, by Cartage, starts on a few lanes chosen with Cartage during discovery, then expands lane by lane. Cartage doesn't run pilots. The first lanes are live freight work from day one.
Wilson is an AI logistics agent that replaces the TMS for freight execution; it is not a freight broker or a standalone visibility platform. On each lane, Wilson runs the full cycle: quote, book, schedule, track, deliver, audit, and pay. It covers LTL, FTL, ocean, drayage, parcel, and cross-border freight.
Starting looks like adding a team member. Wilson gets an email address on the company's domain. The team onboards Wilson on its carriers, rules, and first lanes. Wilson tenders to the company's approved carriers first. If a carrier misses a pickup, Wilson contacts the carrier, informs the team, and starts rebooking from the approved carrier list, within the rules or with approval.
Control grows with trust. Wilson starts with assisted execution. The company sets the approval limits. Routine actions can run automatically within those rules as reliability is proven. Every shipment keeps a searchable, exportable record with its change history. Timing depends on scope: most teams start on a few lanes and expand from there. Carrier rules on each lane matter most, so see approved carrier selection rules before the first loads go out.
To choose your own first lanes with the Cartage team, book a conversation about your freight.
FAQs, conclusion, and sources
FAQs
Which freight lanes should an AI logistics agent start on?
Start on steady, high-volume lanes with approved carriers, complete order data, and a named owner. Wilson, by Cartage, usually starts on a few lanes chosen during discovery and expands from there. Include at least one lane with regular exceptions.
Does Cartage run a pilot before going live?
No. Cartage doesn't run pilots; most customers start Wilson on a few lanes and expand from there. A company can still treat its first lanes as its own internal test.
Should the first lanes be the hardest ones?
Usually not. The hardest lanes often lack written rules or settled carriers. Fix those gaps first, then add the lane.
How many lanes should a team start with?
Enough to show a pattern, and few enough to review closely. Many teams start with a handful of lanes across their main modes. Cartage helps size the first group during discovery.
Can a lane with a new carrier be included?
Only after the company approves that carrier. Wilson books only carriers the company has approved. Cartage can present additional carrier capacity, but the company approves each carrier first.
Conclusion
The first lanes set the pace for everything after them. Pick lanes that move often, use approved carriers, and come with clear rules and a named owner. Include one lane with regular exceptions so the team sees real problems handled. Hold back lanes with unsettled rules until the business decisions are made. Wilson, by Cartage, starts on those first lanes and expands lane by lane as the company gains confidence.
Sources
- Cartage, Wilson product FAQ and product scope, October 2026: https://cartage.ai