Industry · Transport & logistics

AI in logistics. From demand forecasting to dock-to-stock.

We build AI applications that suit the full breadth of logistics: warehousing, distribution, supply chain, last-mile and planning. Custom solutions where the enterprise suite is too heavy and off-the-shelf software is too thin, with the WMS, TMS and ERP integrations that keep your operation running.

DomainWarehouse & DC
DomainSupply chain planning
DomainLast-mile & fulfilment
DomainDocuments & customs
Arrange an introductory call See our work

The Dutch logistics sector in figures.

~14.000
Logistics service providers in the Netherlands
~770.000
Jobs in transport & logistics
~9%
Contribution to Dutch GDP
63%
Of DC managers see AI as a strategic priority

Source: CBS / TLN sector monitor, Logistiek.nl benchmarks.

The enterprise suite does not suit every DC.

If you want to deploy AI seriously in logistics, you quickly run into the same three names: Blue Yonder, o9 Solutions and SAP IBP. Powerful packages, but implementation runs over a long horizon and the price tag suits an operation moving hundreds of millions in goods. For a mid-market 3PL, a specialist DC or a cold-chain operator, that is a suite that never fully gets used, and one that demands more data maturity than a team still being built can deliver.

Standard packages in the WMS and TMS segment do the opposite: they provide operational basics, but demand forecasting, slotting optimisation, computer vision on the dock and document AI for customs remain separate modules that nobody integrates properly. The result is that planners fall back on Excel, an experienced team leader decides by hand which SKU sits at which dock, and the value of your own order history sits gathering dust in a data warehouse.

Meanwhile, the operation changes every month. An e-fulfilment peak, a new range from a shipper, a changed customs regime, a switch of carrier for last-mile: all moments when a static set of rules causes disruption. AI has a real role here, not as a replacement for the people who know the operation, but as a reinforcement of the signals they would otherwise only see a week later on a KPI dashboard.

Custom AI solves this differently: a model trained on your own order volume, connected to the WMS feed you already have, with an interface your planners will recognise. No second system alongside the first, no second source of truth in the data. A layer that sharpens the existing stack and that your own IT team can maintain once we step out after the build.

AI applications by logistics domain.

The use cases differ considerably by domain. A DC has different questions than a planner at a 4PL, and cold chain brings yet other requirements. For each area, we show where we typically build.

Warehouse & DC

The centre of gravity of many logistics operations. AI delivers direct gains in pick productivity, slotting quality and damage control. A well-trained model on your own order history saves a picker metres per order and measurably reduces handling errors. For 3PLs serving multiple shippers from one DC, the extra gain lies in cross-dock matching: linking inbound freight to outbound shipments before the pallets hit the warehouse floor.

Our warehouse module set includes SKU-level slotting optimisation, pick-path sequencing suggestions that take the physical layout into account, computer vision for damage inspection and volumetric measurement at inbound, and a labour planning forecast based on expected inbound volume. For reverse logistics we also build classification models that automatically route incoming returns to restock, refurbishment or destruction, a use case that hardly scales with traditional rules once the range grows.

Without exception, the models run alongside existing WMS systems such as Manhattan, Körber, Boltrics or Centric. We do not write a replacement picking engine. Instead we deliver decision suggestions that your operation either follows directly or reviews via a dashboard, depending on your risk appetite and what the regulations permit.

  • Slotting optimisationHigh-runners closer to the dock, dead stock out of the hot zones.
  • Pick-path suggestionsSequencing models that account for layout and weight.
  • Camera-based damage inspectionComputer vision on the inbound dock, photo archive per pallet.
  • Volume forecast for planningInbound volume predicted, labour planning aligned per shift.
  • Anomaly detection in stockDiscrepancies between WMS stock levels and physical counts are automatically categorised.

Built to the standards that apply to logistics.

AI in a distribution centre touches workforce planning, personal data and, in cold-chain or pharma, tightly regulated product flows. From the first sprint, we work within the relevant frameworks.

GDPR

Privacy by design

Dock camera footage, pick statistics per employee and workforce planning models all involve personal data. We deliver a DPIA, separate training and production data, anonymise wherever possible and document retention periods for each data stream. Camera footage is subject to strict rules: short retention, detection only, and no biometric identification without a valid legal basis.

EU AI Act

Risk classification for workforce AI

AI used in workforce planning and performance measurement may fall under the high-risk category. We map the classification of each model, deliver a model card covering data preparation, training method and known limitations, and ensure human review of decisions affecting individual employees. We build the governance layer, covering versions, datasets and audit logs, from the first sprint, so that a later audit does not require a major catch-up exercise.

CSRD & Scope 3

CO2 reporting from logistics data

For shippers and 3PLs subject to the CSRD, we connect the TMS feed to Scope 3 calculations, per shipment, per lane, per mode of transport. Auditable for the external accountant. See also our page on CSRD reporting software.

NIS2

Cybersecurity for critical infrastructure

Logistics providers within critical supply chains fall under NIS2. We work with segmentation, logging, vulnerability management and an incident plan that aligns with your existing security organisation.

EU MDR & GDP

Cold chain and pharmaceutical logistics

For medical device logistics or pharmaceutical distribution, we support temperature monitoring, batch traceability and the associated GDP requirements. AI-based anomaly detection complements your existing data loggers.

HACCP & food safety

Food-safe flows

For food logistics, we ensure that AI decisions (slotting, picking sequence) are compatible with HACCP zones and allergen segregation. No automated optimisation that compromises food safety.

Seamlessly connected to your logistics stack.

We connect to the systems your operation already runs on. Our AI layer complements them rather than replacing them. WMS, TMS, ERP and EDI flows remain leading.

SAP / Oracle
ERP feed
Manhattan / Körber
WMS integration
Boltrics / Centric
3PL WMS
Transwide / Transporeon
TMS & carrier
FourKites / project44
Real-time visibility
EDIFACT & EANCOM
Shipper messages
Customs / NCTS
T1, AGS, customs API
AWS / Azure ML
Model hosting

No ad hoc integrations, but a repeatable integration layer.

We have standardised the integrations mentioned above across several logistics projects. For a new engagement, we set up the feed with the same abstraction layer, which means fewer bugs at the interfaces, a shorter build phase and a data architecture your own IT team can maintain.

For the more complex cases, such as a heavily customised WMS installation or a legacy integration at AS/400 level, we build a dedicated adapter for each case. This takes extra time in the planning phase, but avoids a fragile point solution later. Read how we approach this for a new custom WMS implementation or a TMS replacement.

On the visibility side, we are seeing FourKites and project44 used more and more as sources for real-time ETAs and exception signals. Our AI layer picks up those signals and links them to operational actions: rescheduling a dock slot before the truck arrives, moving an outbound shipment to the front, or proactively informing a customer via the portal. No more standalone alerts, but actions with context.

From data audit to production AI in clear steps.

An AI project in a logistics operation works differently from a conventional software build. Five phases that are recognisable in most engagements.

01 · Data audit

Data volume and quality

A day on site and a week with your data dump. Outcome: which models are feasible with your own data, and where external sources (weather, price elasticity) add extra value.

02 · Use-case scope

Start with the highest leverage

We select one use case with a clear business case — typically slotting, demand forecasting or document AI — and define it tightly. No sprawl.

03 · Sprint build

Working model in production

Each sprint delivers one working component: model, integration, interface. Planners test alongside us from the outset, and models are trained on real historical data.

04 · Cutover

Running alongside existing processes

First, the model runs alongside the existing process and its predictions are compared. Only once quality is proven do we go live as the decision layer.

05 · MLOps

Ongoing maintenance

Models age. We monitor drift, retrain on new data, adjust for seasonal shifts and keep model versions managed through MLflow. A fixed monthly fee covers maintenance and further development.

Work in the logistics sector.

DC operation · Brabant

Slotting AI at SKU level

Monthly reallocation of pick locations based on SKU-level demand forecasting, integrated with an existing Körber WMS. The model takes into account weight, rotation speed and the physical layout of the pick paths.

Substantial
reduction in pick distance
12.000+
SKUs under management
3PL · Randstad

Document AI for customs

OCR and LLM pipeline for CMR, T1, AWB and certificates. Automatic extraction into the TMS, with human review of exceptions and a feedback loop that lets the model learn from operator corrections.

Significant
time saved per shipment
4 languages
recognised automatically
Cold chain · Nationwide

Temperature anomaly detection

A real-time model that detects deviating temperature patterns before the threshold is exceeded. Logged in line with GDP and linked to an escalation flow to the chain's qualified person.

earlier
alert than classic loggers
3 parties
in the same chain

What the sector writes about AI in logistics.

Logistiek.nl, autumn 2025

"The Dutch logistics sector no longer sees AI as pilot territory — slotting, demand forecasting and document AI are use cases in which mid-market 3PLs are now actively investing."

TLN sector monitor

"The biggest brake on AI adoption in logistics is not the technology, but the integration with existing WMS and TMS installations. Whoever solves that integration reaps the rewards."

Customer conversation, 3PL Randstad

"We rebuilt the document pipeline with custom development, as Blue Yonder proved too heavy for us. The payback period fell within the first financial year."

Answers for the logistics manager starting with AI.

The questions we hear most often from COOs, supply chain managers and DC directors — answered from the practice of our own projects in the logistics sector.

What is the difference between AI in logistics and AI in transport?
AI in transport mainly revolves around vehicles: route optimisation, telematics, predictive maintenance for trucks, ETA models for journeys. AI in logistics is broader and covers the entire goods flow: warehouse optimisation, demand forecasting, stock allocation, cross-docking, reverse logistics, document AI for customs and labour planning. Those focused solely on transport usually have a TMS-oriented question; those in logistics typically have a WMS- and planning-oriented one. We have also written a separate page on AI in transport.
Which quick wins can be delivered within a few sprints?
We consistently see three use cases as a starting point: document AI on incoming freight documents (CMR, AWB, T1), because the business case is clear and the data is already digital; an initial demand forecast for your top 100 SKUs, because historical data is usually sufficient there; and a slotting analysis showing which SKUs are in the wrong place. None of these three requires major infrastructural changes or extensive data preparation. We integrate directly with the WMS or ERP feed you already have, deliver a first working output within a few sprints, and only then decide whether the use case is rolled out more broadly.
Do you replace Blue Yonder, o9 or SAP IBP?
No — for enterprise demand and supply planning at the scale those suites are built for, we refer clients to the right partner. We build custom software for mid-market organisations that find the enterprise suite too heavy, for industry-specific challenges (cold chain, pharma, food with HACCP rules) and for deep integrations between five or more systems that no suite covers out of the box. White-label visibility platforms for 3PLs and customer portals for shippers — with their own branding and their own flow — also fall into this custom segment. We build these from the first sprint.
How do we handle GDPR and the EU AI Act in workforce planning?
AI in workforce planning falls under two regimes at once. For GDPR, we work with data separation, a DPIA and clear retention periods for pick statistics per employee. For the AI Act, we map out the model's risk classification, since workforce planning may fall under high-risk, and we ensure human review of individual decisions, a model card, and transparent explanations for the works council. We do this from sprint 1, not as an afterthought.
How much data do you need for demand forecasting?
It depends on the type of product. For stable SKUs with seasonal patterns, we work with several years of order history at weekly level. For fast-moving e-commerce assortments we need less history, because we incorporate external signals (weather, price elasticity, marketing spend) through demand sensing. For new SKUs we use hierarchical forecasting at product group level until there is enough history on the individual SKU. We work with both classical time-series models (Prophet, NeuralProphet, ARIMA variants) and neural architectures (DeepAR, Temporal Fusion Transformer), depending on data density and the complexity of the assortment.
Computer vision in the warehouse: what is realistic?
Damage inspection at the inbound dock, volumetric measurement for shipping costs, number plate recognition on the site, pallet condition checks and empty location detection. These are all applications that work well in a DC environment with an industrial camera and a trained model. We usually work with PyTorch-based models, hosted on AWS or Azure, and feed the outcomes directly into the WMS so the operator receives immediate feedback. What we advise against: vision systems that make production decisions without a human fallback, because the operational cost of an error is too high there. Warehouse robot coordination (Locus, Geek+, AutoStore) we leave to the robot vendors themselves; we mainly integrate at the data exchange level.
What does an AI project in logistics cost?
A first, sharply defined use case (for example document AI on one type of consignment note, or a demand forecast for your top SKUs) is a compact engagement with a fixed sprint budget. A DC-wide AI layer with multiple models, MLOps and integrations with WMS, TMS and ERP is a larger engagement. We always work with a fixed planning phase in which we establish the scope clearly; only then do we give a concrete price for the complete build, including support during the initial period after go-live. Read how we approach this for custom LLM integrations, which is where much of our document AI comes from.
How do you start in a mid-market DC without a data team?
With one use case and your existing data. We carry out the data audit, deliver the first pipeline and set up a dashboard that lets your operations team see the model outputs for themselves. Only once the use case has proven its value do we expand further. You don't need an in-house data team to get started, though it is useful to have one in mind within a year or two to help plan the next steps. We deliver the models, monitoring and documentation in a way that a mid-market IT department can take over without having to build a full ML stack itself. For the broader context, see our page on enterprise AI implementation.
Which models and frameworks do you use as standard?
For time-series demand forecasting we work with Prophet, NeuralProphet and, where there is enough data, DeepAR or a Temporal Fusion Transformer. For document AI we use OCR engines combined with an LLM layer for extraction, validation and exception detection. For computer vision we work with PyTorch models on an industrial camera feed. The infrastructure runs on AWS or Azure, with MLflow for model versioning and a dedicated monitoring layer for drift detection. No vendor lock-in: you can always move to your own environment.

Talk to us about AI in your logistics operations?

A thirty-minute introductory conversation with your supply chain manager or DC director. We listen, ask questions about your systems and flows, and suggest a first direction for where AI will deliver the most value. No obligation.

Schedule an introductory callEmail fabian.vandijk@appfront.nl

Edit content