AI-supported forecasting introduces the concept of targeted availability, with easily available parts “planned more leanly” and parts more expensive and difficult to get are “reliably secured”. Moreover, the model provides information on mileage and maintenance levels creating the conditions for better forecasting.

The oil pump case

One example of how implementing this AI-supported model in Darmstadt has improved forecasting of spare parts is the case of oil pumps. “While the old method did not identify any demand, the AI model predicted five units – actual consumption was six. With delivery times of around 500 days, this determines whether a vehicle is out of service or remains operational”, DB Cargo explained.

The survey is entirely anonymous and takes up to 10 minutes to complete. Key findings will be shared with our readers in the coming months. Respondents should be directly involved in their company’s IT landscape and/or digital solutions related to combined transport.

The survey addresses the following market segments within the combined transport sector:

The link to the survey can be found here: https://forms.gle/SPYvR7egf6pS8WhB7

Digitalisation still outpaces AI

Across the global sample of terminal operators surveyed, digitalisation remains the top priority, far ahead of artificial intelligence. According to the UserEvidence dataset:

Even among large terminal networks (11+ sites), where AI interest is strongest, digitalisation still comes first: 64% prioritise it. In comparison, 73% mention AI — a rare case where AI rises, but still relies on other foundational projects underway.

This aligns with Tideworks’ long-standing stance: before AI can improve planning or forecasting, the underlying systems must communicate cleanly, consistently, and in real time.

Data friction the real bottleneck

The survey reveals a common challenge that underscores why AI can’t simply be “plug-and-play”:

These figures reinforce a point often raised by Tideworks: AI does not replace the need for high-quality data flows. If terminals still rely heavily on spreadsheets, manual updates, or fragmented reporting — as 44% still do — then AI tools have nothing reliable to work with.

Before algorithms can optimise capacity or predict disruptions, terminals must tighten core digital processes and standardise information sharing.

Operators want practical wins

Asked about priorities for the next 12 months, the industry’s focus is refreshingly pragmatic:

These are operational fundamentals. They reflect daily pressures: meeting throughput targets, controlling costs, dealing with labour shortages, and keeping equipment productive.

AI may help with some of these in the long run. However, today’s operators are still strengthening their foundations: improving their terminal operating systems (TOS), digitising appointments, expanding automation, and streamlining data flows.

In the survey findings, one project manager from Rail Cargo Group noted that improving the TOS, enhancing reporting, and implementing OCR will be critical before unlocking further automation and operational intelligence. This perspective appears repeatedly across the industry data: operators want enhancements they can measure, trust, and deploy incrementally.

Where AI begins to find its place

Even though AI isn’t leading the agenda, it is emerging in specific pockets. For instance:

This suggests that AI adoption correlates with digital readiness. Terminals with strong TOS deployments, automated gate systems, OCR, and dashboards — all widely implemented across the same dataset — offer a better starting point for AI-driven optimisation.

Valuable benefits result from core digitisation

When terminals were asked what they expect to gain from digitisation, the top responses were concrete and immediate:

These are benefits that do not require advanced AI. They come from automated gates, integrated booking systems, modern TOS platforms, standardised data exchanges, and reliable visibility tools — all areas Tideworks has long emphasised in its ecosystem approach.

Why “AI is not the end-all be-all” resonates

The combined message from the research and real-world observations is unmistakable:

AI will play a growing role in the sector — but only once the fundamentals are modernised.

Looking ahead: Build the foundation first

The freight industry is entering a period where digital maturity determines competitive advantage. AI is part of the future, but it is not the foundation. Instead, the foundation looks like this:

This is where operators are investing today — and where vendors like Tideworks continue to focus their development efforts.

AI will eventually sit atop this digital infrastructure. But until the basics are strengthened globally, it’s clear from the numbers that most terminals want technologies that solve today’s real problems, not tomorrow’s hypothetical ones.

One of the core answers KIBA delivers is a new approach to network planning optimisation for combined transport.

The planning pain behind today’s intermodal networks

In many intermodal networks, planning still reflects the constraints of the past. Capacity is often planned per relation, per terminal or per individual train. Planners work with spreadsheets, static capacity tables and a lot of experience. That experience is valuable – but it has limits once you try to coordinate hundreds of relations, dozens of hubs and a constantly changing demand pattern.

At the same time, demand is volatile, and disruptions are part of daily life. Customers book late, cancel or upgrade loads at short notice; infrastructural works and operational delays force last-minute changes. Trying to keep a European-wide network balanced under these conditions, with only manual planning, inevitably leaves potential capacity on the table and makes it harder to keep rail competitive against road.

KIBA in a nutshell

KIBA is a multi-year research and innovation project funded by the German Federal Ministry for Digital Transformation and Government Modernisation. Led by Kombiverkehr KG as combined transport operator and network coordinator, the consortium brings together Deutsche Umschlaggesellschaft Schiene-Straße (DUSS) as terminal operator, VTG as wagon and railcar specialist, INFORM as optimisation and AI software provider, Goethe University Frankfurt and the Technical University of Darmstadt as academic research partners, and KombiConsult as consulting firm for combined transport.

The goal: develop and test a demonstrator for a capacity management and loading optimisation system in combined transport. Conceptually, it consists of three optimisation modules:

This article focuses on the third component: network planning optimisation – deciding which load units should travel on which trains, along which routes and at which times, to make the best possible use of the available network. Further below, we also provide an insight into how network planning optimisation interfaces with train load optimisation.

What network planning optimisation actually does

From an operational perspective, network planning optimisation in KIBA addresses a simple but powerful question:

Given the capacities of trains for their particular routes over the next days and weeks – how should all current and expected load units be routed through the network to achieve the best overall result?

Instead of planning train by train or terminal by terminal, network planning optimisation looks at the entire system. It decides:

The objectives reflect what operators and their customers care about: maximise utilisation and transported volume, minimise transport time and cost, and avoid overloads, congestion and unnecessary re-handlings at hubs.

The data foundation: forecasts meet bookings

KIBA builds on three pillars for the network planning optimisation:

In the KIBA demonstrator, a central controller gathers these inputs, combines them with static network and timetable data, and then triggers the appropriate optimisation runs.

How the optimisation runs in practice

For a combined transport operator, network planning optimisation in KIBA supports two main modes of use.

Interfacing with train loading optimisation

Network planning optimisation in KIBA does not exist in isolation: its decisions can be refined even further at a train and wagon level.

Upstream, the network optimiser works with aggregated capacity – length, weight and slots on each train and segment. Downstream, the train-loading optimiser takes the result for a given train – the list of load units assigned to it – and computes a detailed plan that places each unit on a specific wagon position while satisfying all technical loading rules and safety regulations.

By feeding relevant wagon and loading constraints back into the network level, KIBA ensures that the “perfect” network plan is not one that later turns out to be impossible to load in practice.

What kind of gains are realistic?

Tests with real data show that network-wide optimisation can significantly increase average utilisation of available train paths and wagons, because it aligns expected and actual demand with capacity across the full network rather than train by train.

At the same time, the optimisation model actively penalises unnecessary re-handlings and late deliveries at the destination. That steers flows away from overloaded hubs and fragile routings, helping to reduce congestion and delays.

In disruption scenarios, the ability to recompute routings quickly supports a more controlled reaction: instead of ad-hoc problem-solving, planners see consistent proposals that preserve service levels as far as possible and make transparent where compromises are unavoidable.

In validation with historical data, the optimised plans were nearly identical to real-world decisions made by experienced planners and revealed additional capacity or more robust routings that would be hard to identify manually in a complex network.

For an industry trying to shift more cargo to rail without waiting for new infrastructure, these are exactly the types of gains that matter: more volume, more stability, more transparency – with the assets already available.

Looking ahead: corridor-level potential

So far, KIBA focuses on a single operator’s network. But the logic extends naturally to international corridors and multi-operator contexts. If partners are willing to share at least aggregated capacity and demand information, similar optimisation approaches could support:

For an industry under pressure to deliver more with less, network planning optimisation of the kind explored in KIBA is not just a research curiosity. It is a concrete step toward turning European combined transport networks into living, data-driven systems that respond intelligently to demand and disruption – and make rail a more attractive default choice in logistics.

OCR Gates: Automated data capture and damage detection

At the terminal’s entrance, OCR gates instantly record the details of arriving trucks and trains, while AI automatically detects container damage. The system can identify nearly thirty different types of defects, such as punctured sidewalls, greatly supporting staff in inspecting cargo. Full panoramic images are captured for every train and truck, allowing operators to determine whether any damage occurred before or after arrival.

Safety is further reinforced by the Milestone camera system, which continuously monitors the terminal. This intelligent solution goes beyond traffic recording: it also enhances workplace safety by automatically flagging if someone is not wearing mandatory protective gear, such as a helmet or high-visibility vest.

Digital Twin: Real-time visualisation

Taking digitalisation to the next level, a 3D digital twin provides real-time, graphical visualisation of everything happening across the terminal—day or night, in rain or fog. Beyond visualisation, the system offers powerful analytics: for example, the status of refrigerated containers can be checked instantly, and virtually any operational data can be processed and displayed on demand.

EWG’s development team is also exploring how artificial intelligence can be applied even more extensively in the future. Plans include AI-based predictive and decision-support solutions to boost logistics efficiency further, optimise resource utilisation, and raise operational reliability to an even higher level.

Integrated control with 5G and centralised data

The terminal’s 5G-based crane control system is a European first. Thanks to ultra-low-latency connectivity, cranes can be operated remotely with pinpoint precision and safety, paving the way for future full automation.

A Kafka-based central data platform ties all these technologies together. Every piece of incoming information is collected in one place, where it can be analysed, replayed, and archived in real time. As a result, just three operators can safely and comprehensively oversee the entire 85-hectare site. Facility management is further strengthened by a SCADA system that centrally monitors the power grid, energy use, and key infrastructure such as wastewater pumps.

Planned mobile app for drivers

Looking ahead, EWG is considering the development of a dedicated mobile application to provide drivers with real-time, on-site navigation. The concept envisions a digitally mapped internal road network that would guide trucks precisely to their designated locations while enabling dispatcher–driver communication through the same channel. The goal would be to speed up movements among the 200 parking spaces and introduce additional monitoring capabilities.

These forward-looking technology initiatives highlight the direction in which intermodal logistics is evolving: toward a fully digitalised, integrated 5G-based ecosystem designed to deliver efficiency, safety, and transparency for operators and customers alike.

Extending this use of data to other echelons of the supply chain can yield considerable benefits, points out Van Derrick. “Automatisation in terminals ultimately helps the entire chain, but it would work better if there was a data exchange between all the component parts.”

For example, an AI system that analyses historical dwell times, crane schedules and gate congestion could predict when a container will be available for pickup. That data can then be exchanged with a truck company, enabling more dynamic and optimised time slots.

Data exchange

Nevertheless, data formats used by the terminal may not be compatible with those used by external partners. That means that such data exchanges are often difficult to put into practice. According to a Tideworks survey, 60% of terminals consider integration with external partners and customers as their top data management challenge.

In order to overcome such barriers, data needs to be standardised industry-wide. The TIC4.0 initiative, of which Tideworks is also a participant, aims to achieve that. It brings together companies from both the terminal operator industry and port equipment manufacturers and suppliers to develop such data standards.

All along the supply chain

That could lead to significant progress in the world of supply chains. “Big companies are trying to build a unified supply chain view across the whole freight route”, explains Van Derrick. “Those standards will help with that.”

Better integration across the supply chain can also enable terminals to provide certain premium services. That could mean prioritised availability for customers willing to pay. The terminal could stay ready to handle a shipment for extra payment, and AI can help plan this.

“In the US for example, companies like Walmart and Amazon base their choice of railway on these types of services. And in Europe, companies like Tesco do”, concludes Van Derrick.

However, not everyone is on board with using AI yet. “Although 73% of larger terminal networks identified AI and automation as essential for future competitiveness, only 36% reported actually using AI to collect and analyse operational data”, Tideworks found in an industry survey.

In other words, only half of those who recognise AI’s importance actually use it in practice. RailFreight.com sat down with Chad Van Derrick, Tideworks’ vice president of software product management, at TOC Europe to discuss how to get terminal operators on board.

“People badly understand AI”

A big part of the problem, explains Van Derrick, is a lack of a clear expected outcome. “People badly understand what AI is, it’s a buzzword”, he adds. “They don’t have an outcome in mind, but implementing AI needs to be centred on the outcome we want to achieve.” At the same time, new users may not always have a clear idea of what is needed from them to get the most out of the technology.

It all starts with the data that goes into the system. It needs to be high quality. “An easy place to start is the inventory”, says Van Derrick. “The system needs to reflect what’s on the ground. If that is not accurate, you’re at a loss.” Terminals need people to govern the quality of the data: standardised, normalised and cleansed.

For example, a container has a life cycle and will be in different places, like a train, in the terminal, on the road, coming into the gate, or moving around the yard. Information about the status of the container might be in different systems at different times. Those systems sometimes do not talk to one another, so you need a system to bring the data into one place. “That may not always be an easy task, depending on the maturity of the terminal and the people involved.”

Data visibility and feedback

Then we reach step two: getting visibility into the data. With dashboard, KPIs and reports, you can use your data to inform decision-making. “You’d be surprised how few use it, and if they do, they might use it for a month and then move on to new reports”, says Van Derrick. The key here is consistent, daily use.

“Ideally, you want it as near real-time as you can get. How long are trucks waiting? What is the turn time? How long are we waiting for containers? You want those insights to begin to understand what data is important and the quality of it.”

Once a sense of trust in the machine has been established, we proceed to step three: feeding the machine with data. “It then learns, which will give you a functioning predictive machine with confidence”, says Van Derrick.

It could then tell you ahead of time, for example, expected container dwell times in terminals. “Then you may know that there is a 90% chance that it may be picked up after three days, allowing for you to place it in a strategic location.”

Trust the system

Understanding the basic functionality of AI in terminals is key to getting businesses on board with it. But in addition, says Van Derrick, people need to trust it. “It is important to provide for a feedback system. If an AI-suggested decision did not work out, you need to let it know.”

Without such feedback, AI could even create safety issues. “It really wants to give real-time instructions, which can lead to dangerous situations”, continues Van Derrick. “For example, it could tell a truck driver to suddenly change course, and then the driver runs into another truck. The AI needs training to understand these sorts of things.”

“The goal is not taming technology or AI criticism but making AI productive and building AI competence. This means translating Europe’s economic strength, industrial DNA, and logistical excellence into scalable, market-ready AI systems built in Europe, for Europe, and the global market,” he explained.

Supply chain knowledge greatest EU asset

Herbers highlighted that the operational intelligence of the European industry does not lie in Californian data centres, but “on the shop floor, in complex supply chains, in multimodal hubs”. According to him, this is precisely where the proposed “AI made in Europe” should focus: not as a playground for generative text models, but as a tactical brain of process control and a decision-support system that translates operational complexity into manageable processes.

“Decisive steps are needed from politicians, industry, and companies. Businesses must bring operational AI to the ground – into factories, logistics hubs, and decision-making centres”, he emphasised.

The capacity to adapt, reconfigure, and redesign production or logistics processes is deeply rooted in Europe’s industrial mindset, SME structures, and education system. Places like Aachen, integrating science, industry, and practice, are seen as “epicentres of this movement”.

Despite this potential, Europe faces a “staggering” pace of innovation in China, where “execution trumps debate”. To avoid being left behind, Europe must “invest now in infrastructure, in talent, and in applications”. Relying solely on basic research or navigating regulations like the “AI Act” is insufficient, concluded Herbers.

In fact, studies have shown that AI can reduce total costs by up to 55% and improve train capacity usage by more than 20%. These numbers aren’t just impressive—they’re vital for an industry under pressure to stay competitive. This article takes a closer look at how AI transforms terminal operations, clears up common misconceptions, and explores where its real value lies in daily logistics work.

How AI optimises operations

Think about how supermarkets manage their inventory: AI forecasts demand and predicts when supplies need to be replenished. This helps avoid overstock and empty shelves and ensures that goods are available where and when they’re needed. The same logic applies to terminal operations: Instead of products, we deal with containers, vehicles, and equipment. AI-based decision intelligence helps predict movements, assign resources dynamically, and make better use of space and time.

Another great example of AI in your everyday life is streaming services, such as Netflix or Spotify. Their AI suggests content based on your preferences and viewing history. Similarly, AI in intermodal terminals can make predictive decisions, determining the optimal storage location for swift retrieval or the best train slot for efficient routing. Instead of manually sifting through endless options, AI provides intelligent recommendations.

To fully grasp how AI functions, it is helpful to take a step back and look at a concrete definition. At its core, AI simulates human intelligence, enabling an AI-powered system to perform tasks that traditionally require human judgment. Looking at intermodal container terminal operations, this means that AI algorithms are provided with vast amounts of structured input such as terminal layout, domain-specific knowledge like recurring peak times on certain weekdays, transport schedules or yard space availability. The advantage lies in the speed at which this information can be processed, which no human could ever match. The outcome is an optimised decision for container placements, routing strategies, job combinations and double-cycling or other processes within the terminal.

Of course, the quality of the result depends directly on the quality and completeness of the data provided or on the possibility of cleaning up the raw data and engineering additional data stemming from it. Machine Learning (ML) adds another layer to this process. While often being seen as one phenomenon, ML is actually a subset of AI that uses self-learning algorithms. Instead of relying only on predefined algorithms for decision-making, the system learns from historical data, previous optimisation runs and additional external data sources, such as time of day, weather conditions, or fuel prices. Over time, it identifies recurring patterns and generates new operational knowledge. This knowledge can be reintegrated into the optimisation process, enhancing decision-making further and making the system more adaptive to changing conditions.

A TOS for intermodal needs

INFORM has a long history of applying AI to logistical challenges. Their journey began with operations research and extended to transportation, airports, and, ultimately, the container terminal industry. One of their early AI-driven projects in this industry was with HHLA Container Terminal Burchardkai (CTB), a highly automated terminal, in the early 2000s. This project demonstrated the importance of AI-powered optimisation software for efficient terminal operations.

Building on this experience, INFORM developed a modular and scalable intermodal TOS, recognising the unique needs of intermodal facilities compared to maritime terminals. This TOS is designed for different terminal sizes and automation levels and is suitable for brownfield (upgrading existing sites) and greenfield (newly built) projects. A key advantage is its integration of AI for optimisation across various terminal functions, including crane movements, yard space allocation, and multimodal transport coordination.

Real-world implementations showcase the impact of the company’s AI-powered TOS. At KTL, a brownfield implementation integrated AI-driven optimisation without disrupting ongoing operations, improving turnaround times and reliability through data-driven decision-making. The greenfield deployment at DGT, a major European inland port, established a digital-first approach from the outset, using AI to anticipate congestion, balance workloads, and enhance coordination across different transport modes.

Benefits of AI in intermodal TOS

The integration of AI into an intermodal TOS offers a multitude of benefits across various operational areas. These include:

• Booking module for the creation of transport jobs as well as conflict management, supporting the integration with external booking for consistent information flow
• Accurate billing via automated tracking of all terminal handling
• Simplified gate handling with automated handover or pickup position delegation for truck divers and integration with OCR systems
• Efficient and forward-looking rail control for managing train schedules, as well as loading and unloading, by considering available equipment
• Intelligent crane control to minimise empty travel and optimise handovers and to manage workloads between equipment
• Adaptive yard coordination for optimised allocation of yard positions considering various factors like dwell times and outbound transport proximity
• Dynamic vehicle management for efficient creation and management of internal job orders
• Optimised barge handling for managing incoming and outgoing ship voyages as well as vessel schedules

Practical Implementation of AI: A People-Centric Shift

Successfully bringing AI into intermodal terminal operations isn’t just a matter of deploying new technology—it’s about evolving how terminals work. AI offers clear advantages over traditional manual processes by enabling faster, data-driven decisions and automating routine tasks. This is crucial as terminals face growing volumes and operational complexity. But rather than replacing people, AI shifts the workforce’s focus. Human expertise remains central, with staff taking on more supervisory, analytical, or decision-support roles. This shift requires reskilling and upskilling—think operations planners learning to interpret AI recommendations or gate clerks adapting to smarter scheduling tools.

For AI to truly take root, standardised interfaces and seamless integration into existing systems are essential, but equally important is investing in terminal staff training. INFORM supports this transition by delivering explainable, user-friendly AI solutions while guiding terminals through process adaptation and workforce enablement. Clear, explainable outputs help build trust and understanding, while user-friendly design ensures AI tools complement rather than disrupt workflows. When implemented with both technology and people in mind, AI becomes a powerful tool for sustainable, long-term efficiency.

The future is now

Waiting for others to lead the way is no longer a viable strategy. Digitalisation and automation are already reshaping terminal operations, and those who delay risk falling behind in an increasingly competitive environment. Now is the time to invest in smarter systems, not once the majority has already moved ahead.

However, change doesn’t have to be disruptive. Transformation works best when all stakeholders are involved early and supported throughout the process. With modular solutions, even smaller terminals can take manageable first steps and scale gradually as their needs evolve. This makes scalable optimisation achievable without the need for a complete system overhaul.

The German rail operator has been using camera bridges with AI to do damage checks for a while, it says. “A total of 13 camera bridges are located at the eight largest freight stations in Germany. Freight trains pass through portals that are equipped with high-speed cameras and lighting. An AI analyzes the captured images for damage”, the company says.

Nationwide AI test

“The new AI technology is now being tested nationwide”, DB adds. “It is intended to speed up the provision of freight wagons and at the same time make the work of employees considerably easier.” Earlier, employees had to check each wagon manually, which was a very time-consuming process, the company says.