Volkswagen Group: AI and automated driving set to transform vehicle logistics and yard operations
At Volkswagen Group Logistics, automated driving, AI and data-driven optimisation are reshaping vehicle logistics, helping to improve efficiency, forecasting and yard management across the OEM’s global networks.
The next step change in finished vehicle logistics may lie in rethinking how vehicles move before they even leave the yard, according to Peter Hörndlein, managing director of vehicle logistics, Volkswagen Group Logistics. In his keynote at ALSC Europe 2026, Hörndlein sets out a vision for a more automated, AI-driven logistics network that is designed to tackle growing volatility, labour shortages and capacity constraints.
At the centre of the transformation is a shift towards automated driving within logistics environments. At its port in Emden, Volkswagen is piloting ‘AutoLog’, its infrastructure-led vehicle automation using LiDAR sensors, 5G connectivity and centralised control software to remotely steer vehicles across logistics yards. Unlike autonomous driving for end customers, this approach relies on external intelligence to manage vehicle movements, enabling driverless operations for tasks such as parking, reshuffling and dispatch preparation.
A big part of this pilot scheme, Hörndlein says, is a focus on efficiency and safety. “We perform a lot of tasks like parking, reshuffling cars, getting them out of rows, and making them interact with manually driven cars. We are for sure looking at the safety side of it,” he explains. “The results so far are very promising, because we can really perform all sorts of complex maneuvers and tasks in the test area. And our strong belief is that we can scale it up and roll it out.”
In order to do this, VW Group Logistics needed a standard communication interface to the remote vehicle operating system to give out the task orders, and also needed a standard communication between the vehicle operating system and the vehicle backend of the car itself. For the latter, he says there is already an ISO communication standard in place. “I can really only preach that adherence to that is crucial in order to scale the whole setup, and we are working with our partners in order to help define these communication standards” he says.
When scaling this technology up, Hörndlein believes real value could lie in using it at ports and compounds where many vehicles from different OEMs are loaded and unloaded.
Digital transformation in vehicle logistics: AI, data and automation reshape network design and yard operations
In a panel following the VW Group Logistics keynote, Frank
Schnelle (pictured), executive director of the Association of European Vehicle Logistics
(ECG), Dennis Feddern, senior vice-president, vehicle logistics at Inform, and Jason
Blood, chief commercial officer at Sphere Global join Hörndlein to further discuss embracing data, automation and connected platforms to unlock agility, reduce costs, and enhance service in a rapidly transforming market.
Network design and planning
Intelligent network design is underestimated, despite network complexity being a top challenge in vehicle logistics, according to the panel. Network design must consider alternative routes and volatility, and pre-planning is essential to react to volatility.
Planning cycles vary, and accurate planning is crucial for LSPs. LSPs need timely information on announced and planned cars. Managing the 10% inaccuracy in planning is a struggle. The tendency to contract more capacity than needed impacts overall planning.
Digital collaboration and data transparency
Digital collaboration platforms are vital for information sharing between OEMs and logistics providers. The OEM mindset has shifted, recognising the need to manage their supply chain ecosystem.
Data transparency along the entire process chain holds significant potential. All parties are now more open to sharing data and information, moving away from silo optimisation.
Data governance requires a platform for agreeing on data sets and what should be shared. The data owner remains the data owner, and OEMs, as the contracting party, should manage data flow to ensure timely delivery to suppliers. Sharing data with competitors presents a difficulty. Lifting overall efficiency requires collaboration, not just individual silo optimisation. The group agree that a common ground for sharing siloed information is needed for better-informed decisions. Collaboration is crucial, as all parties are interdependent, and data exchange is crucial, with value to be gained and given by all parties.
Automated yard operations and damage detection
Automated damage detection makes yards proactive.
Enables decisions on repairing damage or sending vehicles to dealers.
Allows for cost analysis and optimized yard operations, such as sidelining vehicles to specific lanes.
Creates a digital vehicle passport to track where damage is first seen, enabling chargebacks and proactive damage prevention.
Shortens the time to allocate claims, which can otherwise take three to six months.
Real-time visibility of vehicles and movements in the yard changes operational decisions.
Key to finding parked cars and accurately locating them for the next stage.
A digital passport can provide vehicle status, including detected damages, to avoid blaming the wrong party.
Maintenance services like battery recharging or tire pressure control can be linked to the digital passport.
Combining with automated driving allows cars to self-drive for services.
Location accuracy and data timeliness are key for automated and driverless vehicle movements.
AI and spatial sensors can track vehicles without RFID tags, providing accurate locations.
Integration with manual systems is essential, especially with multiple manufacturers at different technological levels.
Safety is a major concern when using robots alongside autonomous vehicles in yards.
Good yard knowledge helps determine vehicle dwell time and yard efficiency.
Handover points are critical to prevent operational fragmentation.
The vehicle's journey from plant to dealer can involve many companies and handover points.
Too many handover points and silos of information make damage allocation time-consuming.
A digital passport helps understand the vehicle's route and prevent damage.
Robotics and automated driving
Robotics can change compound design by eliminating the trade-off between parking space utilisation and driver effort. Systems can operate continuously, creating a constantly moving "organism" compound. Proper planning is essential to leverage these systems effectively, linking to demand planning. Robotics can act as a bridging technology for areas without uniform standards or for used cars.
Previous attempts to introduce robotics in vehicle compounds stalled for two main reasons:
Outdoor limitations: Robotic solutions faced limits in open spaces with rain, snow, or uneven ground.
High cost: The investment in a large number of robotic devices was too costly for business cases.
Automated driving, using the vehicle's own tires and wheels, is preferred for plant yards over robotics, as there is less concern regarding pavements, snow, and other elements, and combining robots and automated driven vehicles on the same yard is not advisable due to complexity.
The return on investment for automated driving in a single plant yard operation can be within two years, depending on the volume and the proportion of cars capable of automated steering. Maintenance is not a significant cost, as LIDAR sensors are robust. This reduces human error and allows for tighter parking, increasing space utilisation.
For multi-brand ports and compounds, standardisation of communication is crucial.
An ISO standard for remote vehicle operations (RVO) systems and car interfaces exists and should be adhered to.
Communication standards from back-end operating systems to RVO systems also need standardisation.
Avoiding fragmentation and silo thinking requires stakeholders to engage and agree on standards and investments.
Port operators will not want to invest in different systems for different OEMs.
OEMs must share vehicle data for autonomous driving to be effective.
Demand prediction and capacity planning
Breaking up silos between emptying the plant compound and organising global transport is essential for efficient lead times. Pushing out the wrong cars at the wrong speed makes subsequent handling difficult. OEMs are combining global planning and transportation approaches to ensure plants are emptied quickly while pushing out the right vehicles. Collaboration between OEMs and LSPs (who manage factory compounds) is key. There is significant "sleeping value" due to vehicles resting in the wrong spots of the supply chain.
Used vehicle flows are harder to predict, requiring flexibility of plus/minus 30% in volume and capacity.
An AI-supported solution analyses data (age, specs, driving behaviour, mileage) to better predict where cars will appear.
Helps in collecting cars from dealerships and proactively moving them to points of sale based on predicted likelihood of sale.
This model helps make more efficient use of scarce capacities (yard space, trucking, drivers).
Even an 8-10% improvement in prediction accuracy is a significant step change.
Future outlook (in five years)
A much higher share of cars will have automated driving capabilities, with the first autonomous cars interacting. There will still be a high need for physical transport infrastructure (railway networks, trucking, shipping). A significant increase in the BEV share in finished vehicle logistics is expected.
Progress is being made in shipping (LNG, ammonia investigations). The trucking side needs acceleration to optimise the carbon footprint.
Total transparency in damage detection will be achieved through a complete digital vehicle passport, and claims related to logistics will decrease due to better data and understanding of damage causes.
Vehicle logistics will catch up to general cargo in applying AI, which will help sort out fragmented information from various sources (Excel, email, calls). The industry will be more accustomed to using data for its own purposes.
Transparency and visibility will remain on the agenda, but with improved alignment among stakeholders. The focus will be on ensuring everyone benefits from sharing data, fostering true partnerships.
AI and reinforcement learning at VW Group Logistics
Beyond automation, Hörndlein points to artificial intelligence as a critical enabler across the logistics chain, from improving demand forecasting accuracy to optimising yard operations in real time. Together, these technologies could move the sector beyond incremental gains towards fully optimised, scalable and data-driven operations.
He says that VW is now introducing a reinforcement learning AI supported tool to develop parking strategies to best optimise capacity and manoeuvres, and says it’s something that could be used in future to presort vehicles into batches for their destined markets, further cutting down time for delivery in vehicle logistics.
With the pilot AI-learning project, VW Group Logistics increased its accuracy in demand prediction for used vehicles by 8% “in a relatively short period of time”. He says: “Imagine what that means, that you can apply your capacities and assign your capacities in a much better way.”
