--- title: "AI for logistics and transport: quotes, disputes | Marsanne" description: "AI systems for hauliers and logistics providers with 20 to 500 people: quotes, disputes, client requirements. Every result is checked." url: "https://marsanne.studio/en/industries/logistics/" lang: "en" --- Logistics # AI for logistics and transport, for companies with 20 to 500 people Quotes, disputes, requirements specific to each client. Marsanne puts AI systems into service on the processes that make up the core business of a haulier or logistics provider, and every result they produce is checked by a second system. Updated on 28 September 2026 ## In brief - Who it is for: regional hauliers, freight forwarders, logistics providers and warehouse operators with 20 to 500 people. - What the system produces: documents in each client's format, quotes ready for approval, complete dispute files. - How it is checked: a second system verifies every result against the client's specifications, regulatory constraints and the branch's margin, and a dated record is kept. - How long it takes: a few weeks of development, without touching the TMS or WMS to get started. ## AI in transport and logistics: where companies stand In 2025, 9% of transport and storage companies with 10 or more employees use at least one AI technology, compared with 18% across all industries. This is one of the lowest rates measured by Insee. Across all industries, 54% of companies that do not use it cite a lack of expertise ([Insee, enquête TIC 2025](https://www.insee.fr/fr/statistiques/9025878)). In a company of 20 to 500 people, where margins are thin, value is lost in four places: - **Incoming documents.** CMRs, delivery notes and freight invoices arrive scanned or photographed, and a team re-keys them. - **Quoting.** Some price requests are never quoted, for lack of time, and sorting is based on the availability of the operations planner. - **Proof.** Clients, insurers and customs ask for documents that have to be gathered shipment by shipment. - **Requirements specific to each client.** Tracking formats, documents, business rules: when configuration costs too much, the account is turned down. Demonstration scenario ## Take on the clients whose requirements used to put you off The company described is fictitious. This scenario shows how we approach a problem; it does not report a result achieved at a client. ![A chain with one blue link at the centre connecting the two parts.](https://marsanne.studio/_astro/logistique-maillon.DE2Rof7H_G9Fw5.webp) ### The situation A logistics provider: two warehouses, one hundred and twenty people. A distributor offers it its regional flows, on condition that it works to its rules: dispatch advice notes in its format, labels to its standard, a daily delivery report, and returns handled according to its procedure. Configuring all this in the WMS takes the vendor weeks, then rework each time the specifications change. The provider turns the business down, or accepts it and loses its margin. ### What the company already has The distributor’s specifications and a set of its actual documents. The daily WMS exports: orders picked, shipments, returns. The team’s knowledge of what each client really requires. ### What the system produces The system reads the WMS exports and produces, alongside it, the documents in the distributor’s format: dispatch advice notes, labels, the daily report, and returns handling according to its procedure. The WMS does not change. When the specifications change, it is this layer that is updated. ### What is checked A second system, separate from the one that produced the documents, checks them before dispatch: - each document complies with the format and mandatory fields of the specifications; - the quantities stated match the quantities actually picked; - the deadlines of the returns procedure are met, or the discrepancy is flagged. A non-compliant document is not sent: it comes back with the rule it breaches. The record stays attached to the shipment. The principle is explained in more detail on the [home page](https://marsanne.studio/en/#verification). ### What the system does not do It does not negotiate the commercial terms of the contract and does not replace the WMS. It does not modify any operational data. ### How the result is measured The revenue generated from the accounts the company used to turn down. ## Quotes, disputes and client tracking: three other uses of AI in logistics | | Quoting | Disputes and damage claims | Client tracking | | --- | --- | --- | --- | | Today | The request arrives by email, in free text. The planner finds the rate grids, checks the constraints, calculates and formats the quote. | Gathering the CMR and its reservations, photos, correspondence and the contract takes hours. Some disputes are paid without being investigated. | The customer service team spends part of its day answering “where is my delivery?” Each reply requires cross-checking two tools, sometimes calling the driver. | | The system produces | The parameters extracted from the request, the rate grids and subcontracted freight applied, a quote ready for approval. | The complete shipment file, its timeline and the argument based on the applicable terms. | The reply, with the latest tracking event and the estimated time of arrival. Shipments that are going to miss their slot are flagged before the client calls. | | What is checked | Hard constraints (ADR, tail lift, height, time slots) and the branch’s minimum margin. A quote outside the limits waits for a decision. | Each argument cites the clause it relies on. Missing documents are named. | No estimate is given without its level of uncertainty. Estimates are compared with actual outcomes, and the gap is tracked. | | Data required | Twelve months of quotes with their outcome, rate grids, subcontracted freight contracts. | Transport documents, correspondence, general terms and conditions and client contracts. | Tracking events and delay history. Continuous access to positions is connected once the value has been demonstrated. | ## What Marsanne does not do in logistics - **Real-time vehicle tracking.** Telematics solutions on the market do this. - **Generic administrative tasks.** Expense reports, meeting minutes: off-the-shelf tools are enough. - **Software that already exists.** When a TMS, a WMS or a route planning tool covers the need, we say so and we name which one. To gauge the expected return of a project before launching it, see [Comment mesurer le retour sur investissement de l’IA](https://marsanne.studio/en/insights/how-to-measure-ai-return-on-investment). ## Frequently asked questions **Do we need to connect our TMS or WMS?** Not to get started. The system works alongside them, on exports and on the documents you receive: price requests, CMRs, contracts. A few people use it in real conditions, and the question of integration comes up once the value has been demonstrated. **Our documents are photographed by drivers. Can the system read them?** Yes, in their actual condition: crooked photos, stamps, handwritten notes. A field read with too much uncertainty is flagged rather than guessed. **How long before the first result?** Development generally takes a few weeks. Users benefit as soon as they start using it. **What happens to our data?** Once the solution is in service, Marsanne no longer has access to your data. You remain its sole owner. **Will our teams need training?** No. The system is designed around the actual work of the team that uses it, and requires no training. ## Articles — Logistics - [Global stock: placing goods as close as possible to forecast sales](https://marsanne.studio/en/insights/global-stock-allocation-forecasts-ai) 8 October 2026 - [Customs classification: proposing the right tariff heading, with sources](https://marsanne.studio/en/insights/customs-classification-ai) 4 October 2026 - [Transport disputes: handling each claim with its supporting documents](https://marsanne.studio/en/insights/transport-disputes-ai-claim-file-documents) 28 September 2026 ## In the press The latest articles from the studio's AI watch on AI in this industry, summarised in English with their source. - [Amazon brings its robotics revolution to the Oise](https://marsanne.studio/en/ai-watch/amazon-brings-its-robotics-revolution-to-the-6ec809) Amazon opens a robotised distribution centre incorporating AI in Beauvais, with 1,000 jobs targeted. La Tribune · 9 Oct 2026 · France - [Taxi company G7 turns to S3NS's ultra-secure cloud to develop new AI uses, from voice assistance for drivers to customer review analysis](https://marsanne.studio/en/ai-watch/taxi-company-g7-turns-to-s3ns-s-ultra-secure-cloud-to-develop-new-ai-a63c39) G7 is relying on the S3NS sovereign cloud to deploy several AI uses, including a voice agent for drivers by the end of 2026. L'Usine Digitale · 8 Oct 2026 · France - [G7 taxis turn to S3NS to deploy secure AI use cases](https://marsanne.studio/en/ai-watch/g7-taxis-turn-to-s3ns-to-deploy-secure-ai-use-1ffbc5) G7 is testing four AI use cases (assistant, voice agent, fleet monitoring, review analysis) on the S3NS secure cloud. L'Usine Digitale · 8 Oct 2026 · France - [At RATP, AI deployment runs into social dialogue](https://marsanne.studio/en/ai-watch/at-ratp-ai-deployment-runs-into-social-d9abdb) The CFE-CGC union has filed a ‘social alarm’ at RATP, criticising a lack of consultation on several AI projects already under way. Solutions Numériques · 6 Oct 2026 · France - [AI at RATP: the social alert centres on consulting the CSE](https://marsanne.studio/en/ai-watch/ai-at-ratp-the-social-alert-centres-on-consulting-the-4e0cc2) The CFE-CGC Groupe RATP accuses management of failing to consult the CSE on the rollout of AI, against a backdrop of case law favourable to elected representatives. ActuIA · 5 Oct 2026 · France [All 38 articles on this industry in the AI watch →](https://marsanne.studio/en/ai-watch?secteur=logistique#base) ## Find out where AI can help in your operations The executive diagnostic takes ten minutes. It compares your answers with situations we encounter in logistics and gives you concrete leads. [Start the diagnostic](https://marsanne.studio/en/diagnostic)