BMW Group to deploy humanoid robots in production in Germany for the first time

Munich. The BMW Group is consistently advancing the
digitalisation and use of artificial intelligence in production. A key
element in this effort is “Physical AI”, which combines digital
artificial intelligence (AI) with real machines and robots. This
enables intelligent systems such as humanoid robots to be integrated
into real-world production processes.

For the first time, the BMW Group is now bringing Physical AI to
Europe and launching a pilot project with humanoid robots at the
Leipzig plant. The project aims to integrate humanoid robotics into
existing series production of cars and to explore further applications
in the production of batteries and components.

“Digitalisation improves the competitiveness of our production – here
in Europe and worldwide. The symbiosis of engineering expertise and
artificial intelligence opens up entirely new possibilities in
production,” said Milan Nedeljković, Member of the Board of Management
of BMW AG, Production.

Last year, the BMW Group successfully implemented a pilot project
with humanoid robots at its Spartanburg plant in the United States.
The insights gained from this project are being leveraged to further
develop and scale Physical AI applications.

Unified IT and data model in the production system

Artificial intelligence is already an integral part of the BMW
Group’s production system. From the virtual factory with digital twins
and AI‑enabled quality controls to intralogistics with autonomous
transport solutions, intelligent systems are used in almost all
production steps.

A prerequisite for the effective use of artificial intelligence in
production is a unified IT and data model across the entire production
system. The BMW Group has consistently transformed isolated data silos
into a unified data platform in its production system meaning that all
data is consistent, standardised, and available at all times. This
enables digital AI agents to take on increasingly challenging tasks
autonomously and in complex environments while continuously learning
and becoming available for additional areas of application. The
introduction of intelligent and autonomous decision‑making agents
marks a paradigm shift in production. In combination with robots,
these digital AI agents are what make up Physical AI.

“Our aim is to be a technology leader and to integrate new
technologies into production at an early stage. Pilot projects help us
to test and further develop the use of Physical AI – that is,
AI‑enabled robots capable of learning – under real-world industrial
conditions,” said Michael Nikolaides, Senior Vice President Production
Network, Supply Chain Management at BMW Group.

Digitalisation and artificial intelligence are core elements of BMW
iFACTORY and are the foundation for future‑proof, flexible, and
competitive production.

Humanoid robotics complements existing automation

The BMW Group is strategically expanding its automation portfolio to
include Physical AI and humanoid robotics. Humanoid robots are viewed
as a value‑adding complement to existing automation. Particularly,
they demonstrate potential in monotonous, ergonomically demanding, or
safety‑critical tasks. The goal is to relieve employees and further
improve working conditions.

Center of Competence for Physical AI in Production
consolidates expertise

Its consistent data platform in production makes the BMW Group an
attractive partner for technology companies seeking to test the
possibilities of Physical AI – particularly humanoid robotics – in an
industrial environment under real-world conditions. By setting up the
new “Center of Competence for Physical AI in Production”, the BMW
Group is taking the next step in consolidating its expertise and
ensuring that the full range of knowledge can be leveraged across the organisation.

The BMW Group is following a clearly structured approach. Technology
partners are evaluated according to defined maturity and
industrialisation criteria and tested in pilot projects under
real-world production conditions. Following a theoretical assessment,
an evaluation is carried out at the manufacturer’s site in the
laboratory using real use cases from the BMW production system to test
integration capability. If this phase is successful, it is followed by
an initial test deployment under real-world production conditions at a
BMW Group plant and subsequently the actual pilot phase.

First pilot project with humanoid robots in Europe

In collaboration with Hexagon, a long‑standing, established partner
of the BMW Group in the field of sensor technology and software, the
first pilot project in Europe is now underway. Hexagon’s
organisational unit located in Zürich, Hexagon Robotics, specialises
in Physical AI and presented its first humanoid robot, AEON, in June
2025. Following an initial theoretical evaluation phase and successful
laboratory tests, there was an initial test deployment at
BMW Group Plant Leipzig in December 2025. A further test deployment is
planned from April 2026 to ensure full integration for the actual
pilot phase starting in summer 2026.

The deployment in Leipzig is focusing on testing a multifunctional
application of the robot. It is based on AEON’s design, whose
human‑like body allows a wide range of hand and gripper elements or
scanning tools to be flexibly attached and enables dynamic use on
wheels. During testing and later in the pilot phase, the robot will be
used in the assembly of high‑voltage batteries and in component manufacturing.

Successful pilot at BMW Group Plant Spartanburg provides key
insights for use of humanoid robots in production

The world’s first deployment of humanoid robots at a BMW Group plant
took place at the Spartanburg plant in the United States in 2025, in
collaboration with the technology company Figure AI. The results
demonstrated that Physical AI can deliver measurable added value under
real-world conditions. Within ten months, the robot Figure 02
supported the production of more than 30,000 BMW X3, working ten-hour
shifts daily from Monday to Friday. Figure 02 handled the precise
removal and positioning of sheet metal parts for the welding process –
a task that is particularly demanding in terms of speed and accuracy
while also being physically exhausting. In total, it moved more than
90,000 components and covered approximately 1.2 million steps in
around 1,250 operating hours.

The pilot confirmed that humanoid robots can safely perform precise,
repetitive work steps – such as positioning components with millimetre
precision – and provided important insights for the further deployment
of Physical AI in production.

During the initial test phases with Figure 02, it was essential to
involve all areas of the production IT infrastructure, occupational
safety, production process management, and shop floor logistics at an
early stage.

One key finding was that the transition from the laboratory to the
actual production environment was faster than expected. Motion
sequences trained in the laboratory could be quickly transferred into
stable shift operation. To ensure smooth coexistence with existing
systems, integration into the BMW Smart Robotics ecosystem was
implemented via standardised interfaces.

The body shop was deliberately selected for the test phase in
Spartanburg, as it already has a high degree of automation. BMW Group
employees in this area have extensive experience in integrating new
technologies and processes. For example, material supply on the line
is already carried out almost exclusively via automated smart
transport robots (STR). The project team’s early communication ensured
transparency from the outset and promoted acceptance. The deployment
of humanoid robots was met with great interest among employees and
quickly became a natural part of everyday work during the course of
the project.

The BMW Group and Figure are currently evaluating additional use
cases for deploying the Figure 03 robot.

Additional quotes for use:

Michael Nikolaides, Senior Vice President Production Network,
Supply Chain Management at BMW Group:

“Our aim is to be a technology leader and to integrate new
technologies into production at an early stage. Pilot projects help us
to test and further develop the use of Physical AI – that is,
AI‑enabled robots capable of learning – under real-world industrial
conditions. The successful first deployment of humanoid robots at our
BMW Group plant in Spartanburg in the USA proves that a humanoid robot
can function not only under controlled laboratory conditions but also
in an existing automotive manufacturing environment.”

Michael Ströbel, Head of Process Management and
Digitalisation, Order to Delivery at BMW Group:

“We are delighted to deploy a humanoid robot for the first time in a
pilot project at a plant in Germany. Following evaluation by our
Center of Competence for Physical AI in Production, tests were carried
out in the laboratory and at Plant Leipzig at the end of last year.
This year, our focus is on step‑by‑step integration into our
production system to explore a wide range of applications. The
emphasis is on researching multifunctional use of the robot in various
production areas such as battery manufacturing for energy modules and
component production for exterior parts. With Hexagon, we have found a
proven long‑standing partner with a highly innovative approach to
humanoid robotics for this project.”

Felix Haeckel, Team Lead CoC Physical AI for Production:

“At our new Center of Competence for Physical AI in Production, we
are pooling our expertise to make knowledge on AI and robotics widely
usable within the company. In recent years, we have built up an
international team of experts that, in addition to in‑house research
and programming, is dedicated to the gradual integration of AI into
the existing production system. At the same time, our team in Munich
is driving its own robotics research to set up, support, and further
develop pilot projects in the field of Physical AI at our plants.”

Arnaud Robert, President of Hexagon Robotics:

“We are very pleased to be working with the BMW Group to advance the
use of humanoid robots in real‑world environments.”