Audi and Agile Robots test AI-powered mobile humanoids for parts sequencing in Ingolstadt
Four partners are trialling mobile humanoid robots that pick and sequence parts at Audi's Ingolstadt plant, where around one million picks happen every day.
- around one million
- more than 20,000
- over 650
What Happened
Agile Robots, Audi, Dassault Systèmes and NVIDIA announced at NVIDIA GTC Berlin a proof-of-concept to automate parts sequencing at Audi's Ingolstadt plant using mobile humanoids. At Audi, demand-driven material supply relies on highly complex sequencing operations, with around one million picks a day ensuring that the right parts reach the right workstation in the right order for assembly. In the Production Lab, Audi's innovation space for testing future production and logistics technologies, mobile humanoids from Agile Robots pick parts from material containers and place them in the correct sequence for final assembly of Audi premium vehicles. Audi Supply Chain experts contribute their sequencing expertise and define the operational requirements for testing the use case under real-world conditions.
around one million
Daily picks that supply the right parts to the right assembly workstation in the right order
The project is already underway: before any robot is deployed within Audi's smart intralogistics environment, a virtual twin of the logistics environment is being created using Dassault Systèmes' DELMIA application. Feasibility, timings, ergonomics and safety for the robot's task are tested within that virtual twin on Dassault Systèmes' 3DEXPERIENCE platform, leveraging the company's Industry World Models and NVIDIA Omniverse libraries. In the future, LEO, Dassault Systèmes' AI-powered Virtual Companion, will assist the engineers and accelerate the process from months to days. Munich-based Agile Robots, which already has more than 20,000 robotic solutions deployed with customers worldwide, builds the mobile humanoids to combine humanoid reach and dexterity with the robustness of a mobile base, and they work at human height so they slot into Audi's existing environment.
- Months
- Reduced to days
- Handle high-variant or semi-rigid parts — one of automation's most persistent challenges — using dual-arm manipulation for large and flexible parts.
- Identify, grasp and manipulate a wide variety of components with AI-powered perception, even when stored in mixed containers or presented in varying positions and packaging.
- Work at human height so they fit into Audi's existing environment.
- Combine humanoid reach and dexterity with the robustness of a mobile base.
- Take over complex, repetitive sequencing tasks so employees can focus on higher-value activities.
“The future of manufacturing will be defined by how successfully we combine industrial expertise with artificial intelligence. For decades, picking within the sequencing process has remained one of the toughest automation challenges in automotive logistics. Together with our partners, we are now exploring how AI-powered robotics can help solve this challenge and unlock a new level of productivity in our Audi factories.”
“Humanoid robots are advancing automation beyond fixed, repetitive tasks to the complex, variable work at the heart of modern manufacturing. Audi and Agile Robots are using Dassault Systèmes DELMIA and NVIDIA Omniverse physical AI libraries to simulate and validate mobile humanoids for parts sequencing, tackling one of automotive logistics' toughest workflows with minimal changes to existing operations.”
Why this matters
Audi says automating complex, repetitive sequencing tasks could reduce manual workload, raise efficiency and free employees to focus on higher-value activities — a shift that directly touches logistics workers at its Ingolstadt plant. The proof-of-concept pairs Audi's sequencing expertise with robotics from Agile Robots and simulation software from Dassault Systèmes and NVIDIA, showing how Physical AI is being tested for factory work that traditional robotics could not handle.
Terms in This Story
- Physical AI
- Artificial intelligence built to perceive and act in the physical world, for example by controlling robots, rather than only processing digital data.
- Virtual twin
- A software replica of a real environment or system that engineers use to test and validate how something will work before it is deployed for real.
- Parts sequencing
- Arranging components in the exact order they are needed so that the right part reaches the right workstation at the right time for assembly.
- Intralogistics
- The movement, storage and supply of materials inside a single facility such as a factory or warehouse.
Summarised from the linked release; details can be imperfect — always verify against the original source.