The alert is not the fix: what should happen after a freight exception appears
Key Takeaways
- Freight exception management is the work of resolving a shipment problem after it's detected: contacting the carrier, updating internal teams and customers, and rebooking or escalating when needed. Real-time visibility detects the exception; exception management fixes it.
- A complete freight exception management workflow has 5 steps: detect, confirm with the carrier, notify internal teams, notify the customer, and rebook or escalate.
- The 4 most common freight exceptions an AI agent can handle are missed pickups, carrier cancellations, in-transit delays and missing proof of delivery.
- Delays are costly: ATRI found truck drivers were detained at 39.3% of stops in 2023, costing an estimated $3.6 billion in direct expenses and $11.5 billion in lost productivity.
- Wilson, by Cartage, is an AI agent for freight exception management that executes the response within guardrails: approved carriers only, approval rules configurable per workflow, escalation for judgment calls, and a change history for every shipment.
What is freight exception management?
At 9:40 a.m., a distributor's logistics team gets an alert: a truckload to its largest customer missed its pickup window. The alert took seconds. The fix takes the rest of the morning: calling the carrier, telling sales, warning the customer's receiving team and finding another approved carrier for the afternoon.
Freight exception management is the process of detecting a shipment problem, such as a missed pickup, a carrier cancellation or a delay, and executing the steps required to resolve it, so the freight and everyone depending on it stay on track.
The need is widespread. In a March 2025 survey of 240 supply chain leaders, BCI Global found that only 35% of companies have appropriate, detailed visibility across their supply chain. And when problems aren't resolved quickly, costs pile up: ATRI reported that time lost to truck detention exceeded 135 million hours in 2023.
The benefits of effective freight exception management:
- Fewer missed commitments: problems get resolved before they reach the customer's receiving dock or the production line.
- Faster response: the carrier is contacted and stakeholders are informed immediately, not when someone has time.
- Less manual work: coordinators stop spending their day on calls and status emails.
- Better customer communication: customers hear about changes before they ask.
Real-time visibility vs. exception management: why the alert is not the fix
Real-time visibility tells a logistics team that something went wrong. Exception management is the work that fixes it. Most teams have the first and still do the second by hand.
| Real-time visibility | Exception management | |
|---|---|---|
| Question it answers | "Is something wrong with this shipment?" | "What do we do about it, and who does it?" |
| Output | An alert, an ETA change, a status update | A confirmed new plan, informed stakeholders, a rebooked load |
| Typical work involved | Tracking, monitoring, alerting | Carrier contact, internal and customer notifications, rebooking, escalation |
| Who usually does it | Software | Often the logistics team, by phone and email |
| With an AI agent | Ongoing monitoring flags the exception | The AI agent executes the response within configured rules |
Real-time visibility is necessary, but it's step one of five. When alerts increase and the team's capacity doesn't, the alerts become a queue. For a closer look at tracking data, see how real-time shipment tracking works.
The freight exception management workflow, step by step
Every freight exception follows the same 5-step workflow. For each step, it helps to be clear about the trigger, what the AI agent does and when a person steps in.
| Step | Trigger | What the AI agent does | When a person steps in |
|---|---|---|---|
| 1. Detect | A missed milestone, an ETA change or a carrier message | Ongoing monitoring flags the exception using the best source for each load: carrier API, GPS/ELD data, the carrier's website by PRO number, or email and direct outreach | Rarely. Detection is automatic |
| 2. Confirm with the carrier | An exception is flagged | Contacts the carrier, requests status and a revised time, and follows up until it gets an answer | If the carrier can't be reached within the set window |
| 3. Notify internal teams | The new situation is confirmed | Sends updates to logistics, sales and operations through Slack or Microsoft Teams | Never required, but the team can respond in the same thread |
| 4. Notify the customer | A delivery commitment is affected | Updates the consignee by email or SMS, as configured | For high-priority customers, if the company prefers a personal call |
| 5. Rebook or escalate | The original carrier can't recover the load | Initiates rebooking from the approved carrier list | When rebooking exceeds a cost or time threshold, or the situation falls outside the rules |
The value of automation comes from running all 5 steps consistently, on every exception, whether it's the first of the day or the fifteenth. More on each step: An alert is not a resolution.
How an AI agent handles the four most common freight exceptions
The 4 most common freight exceptions an AI agent can manage are missed pickups, carrier cancellations, in-transit delays and missing proof of delivery. Each has a clear trigger and a clear expected response.
| Exception | The problem | Trigger signal | What the AI agent does | When a person steps in |
|---|---|---|---|---|
| Missed pickup | Freight is still at the dock, and downstream appointments are at risk | Pickup window passes without confirmation | Contacts the carrier, confirms a new pickup time, notifies internal teams and initiates rebooking if the carrier can't recover | Rebooking above the approval threshold |
| Carrier cancellation | The load has no truck | Carrier cancels by email or API | Notifies internal teams and initiates rebooking from the approved carrier list, tendering the company's own carriers first | New cost above the threshold, or no approved carrier available |
| In-transit delay | The delivery window is at risk | ETA change or late status update | Gets a revised ETA from the carrier and updates the internal team and the consignee | Delays beyond a defined window, or a high-priority customer |
| Missing proof of delivery | Invoicing and dispute resolution are blocked | Delivery confirmed, but no POD received | Follows up with the carrier until the POD arrives, files it with the shipment and notifies the team | POD still missing after the set window |
Exceptions that require judgment, such as damaged freight, claims or carrier disputes, stay with the logistics team, and the AI agent handles the notifications around them. For more on POD follow-up, see The shipment isn't done until the POD is filed.
A missed pickup from alert to customer notice, step by step
The table below follows one missed pickup through the full response, with the time of each step and the record that proves it. This example is hypothetical and does not describe a real customer. It shows what a team should be able to check in the shipment record afterward: every step has a time, a source, and an owner.
| Time | Step | What the AI agent did | Record that proves it |
|---|---|---|---|
| 8:00 a.m. | Pickup window closes | Flags the load: no pickup confirmation | Status history on the load |
| 8:10 a.m. | Carrier contact | Emails and then calls the carrier's dispatch: "Load 3317, pickup was due 7–8 a.m. at the Joliet plant. Is the truck coming, and when?" | Sent email and call log on the load |
| 8:35 a.m. | Carrier answer | Dispatch replies that the truck broke down and cannot arrive until tomorrow | Carrier reply saved to the load |
| 8:40 a.m. | Assessment | Compares tomorrow's pickup with the customer's delivery date: it would miss | Note on the load with both dates |
| 8:45 a.m. | Internal notice | Tells the logistics team in Microsoft Teams: original carrier cannot recover, rebooking started | Message on the load |
| 9:15 a.m. | Rebooking | Gets quotes from two approved backup carriers; one can pick up at 1 p.m. for $160 more | Quotes saved to the load |
| 9:20 a.m. | Approval | $160 is above the coordinator's limit, so it goes to the logistics manager, who approves | Approval with name and time |
| 9:30 a.m. | Cancel and book | Cancels the first carrier in writing, tenders the backup, receives acceptance and the signed rate confirmation | Cancellation, acceptance, and rate confirmation on the load |
| 9:40 a.m. | Customer notice | Emails the customer: pickup moved to 1 p.m., delivery date unchanged | Sent email on the load |
| 1:05 p.m. | Pickup | Backup carrier picks up; status updated | Pickup confirmation on the load |
The point of the table is verification. After the load moves, anyone on the team can open the shipment record and see each step, its time, and who approved the extra cost. Wilson keeps searchable, exportable shipment records with change history, so a buyer can ask to see this kind of record for a test load during a demo.
Control and autonomy: the guardrails behind AI exception management
AI exception management works when automation and control come together: the AI agent executes routine responses automatically within approval rules the company defines, and routes everything else to a person.
Wilson, by Cartage, is an AI agent for freight exception management that works within five guardrails:
Approved carriers only. Wilson initiates rebooking from the company's approved carrier list. It tenders the company's own carriers first and books only carriers the company has approved; Cartage can introduce additional carrier capacity, but the company approves any carrier before Wilson uses it.
Configurable approvals. Human approval isn't hardcoded. Each company decides, workflow by workflow, which steps run automatically and which wait for sign-off, for example rebooking above a cost threshold.
Escalation for judgment calls. Anything outside the rules, such as claims, disputes or unusual requests, goes to the logistics team.
A record of every change. Wilson keeps searchable, exportable records of every shipment, including change history, so the team can see what happened and why.
Automation that grows with trust. Wilson starts with assisted execution and moves more work to automatic execution as reliability is proven on each workflow.
How Wilson monitors and communicates. Wilson provides ongoing shipment monitoring using the best source for each load: carrier API, GPS/ELD data where available, the carrier's website by PRO number, and email or direct outreach when none of those exist. It communicates by phone, email, Slack, Microsoft Teams and SMS, and updates the company's team, the shipper and the consignee at booking, pickup, delays, delivery and POD filing. For the full communication model, see freight communication automation.
How it fits. Orders come from an ERP, spreadsheet, email or scheduled file drop. Wilson replaces the TMS for freight execution and connects to the company's ERP. A company without a TMS does not need to implement one first, and a company with one can move freight execution to Wilson lane by lane. It is not a standalone visibility platform; exception management is part of its broader freight coordination, from quote to delivery. Timing depends on scope: most teams start on a few lanes and expand from there. For more on the category, see What is an AI logistics agent?.
FAQs
- What is freight exception management? Freight exception management is the process of detecting shipment problems, such as missed pickups, cancellations, delays or missing PODs, and executing the steps to resolve them: contacting the carrier, notifying stakeholders and rebooking or escalating.
- How does AI exception management work? An AI agent monitors shipments, detects an exception, contacts the carrier, notifies internal teams and customers, and initiates rebooking from the approved carrier list when needed. Wilson, by Cartage, runs this 5-step workflow within approval rules the company configures.
- What's the difference between real-time visibility and exception management? Real-time visibility detects that something went wrong. Exception management resolves it. Visibility produces an alert, and exception management produces a new plan, informed stakeholders and, when needed, a rebooked load.
- What are the benefits of automating freight exception management? Faster response, fewer missed customer commitments, less manual carrier follow-up and earlier customer communication.
- Which freight exceptions can an AI agent handle? The most common are missed pickups, carrier cancellations, in-transit delays and missing proof of delivery. Claims, damaged freight and disputes stay with the logistics team.
- How does a company stay in control of an AI agent? Through guardrails: approved carriers only, approvals configurable per workflow, escalation for anything outside the rules and a change history for every shipment. Wilson starts with assisted execution and expands automation as reliability is proven.
- Does AI exception management require a TMS? Not with Wilson. Wilson replaces the TMS for freight execution and takes orders from an ERP feed, spreadsheet, email or scheduled file drop. A company with a TMS can move freight execution to Wilson lane by lane.
Conclusion
An alert tells a logistics team that something went wrong. It doesn't call the carrier, update the customer or find another truck. That's the work of freight exception management, and for many manufacturers and distributors it still happens by hand, one exception at a time. Wilson, by Cartage, is an AI agent that executes the full 5-step exception workflow, within guardrails the company sets, so every exception gets the same fast response. The practical first step is to pick the exception that costs the most, usually missed pickups or carrier cancellations, and start on a few lanes, expanding from there. Cartage helps identify and map that workflow during discovery.