Top 10 most cutting-edge German startups at Automatica 2022 that develop AI for robotics

The first robot ever to be used in the industry was a 6 degree of freedom (DOF) programmable robotic arm called Unimate (Unimate is a combination of the words "universal" and "automation") (Fig. 1). It was the first industrial robot that worked on the General Motors assembly line for stacking hot metal (Fig. 2). Subsequently, other automakers licensed the technology to aid with their production. [1][2]

Fig. 1.1: Unimate robot

Fig. 1.2: Unimate robot stacking hot metal

The advent of robots reshaped the automotive industry, increasing the efficiency and quality of production lines. Fast forward to today, robots are used not just in the automotive industry but in a wide range of domains such as healthcare, agriculture, manufacturing, military, and food preparation, and for a variety of tasks such as machining, painting, welding, assembly, palletizing, inspection and even 3D printing. [3][4]

Conventional robotic systems that are used to perform tasks such as pick-and-place (Fig. 2), welding, assembly, etc. in the industry assume that the spatial location and geometry of the workpieces are already known. In such cases, the robots follow a pre-determined trajectory, and the robots are limited to operating only in structured environments. This inability to navigate unstructured environments is an inherent limitation of conventional robots. As an example, traditional robots are incapable of maneuvering unexpected events on their path of travel. If an obstacle appears while the robot is moving, it simply halts in order to avoid collision rather than finding an alternative route to reach its target. [5]

Artificial intelligence (A.I.) promises to equip robots with the ability to navigate and manipulate a complex and unpredictable environment. The number of industries that utilize A.I. is increasing at a tremendous scale. This is evident by the emerging number of startups that are starting to utilize approaches like machine learning, language processing, planning, etc. to solve problems that have previously been extremely hard to solve. This is particularly true for robotics where A.I. is making a serious impact and is en route to disrupt the industry. For instance, through machine learning, robots can now handle previously unseen objects with sufficient accuracy to be deployable in real-world applications. Additionally, A.I. enables the robot to maneuver through its path by building a navigational map in real-time. A.I. is also used in the manufacturing industry to find the correct timespan for holistic maintenance of robots, thereby preventing breakdown and the associated repair costs. The A.I. revolution will enable robots to handle highly complex cognitive tasks and the trend will undergo widespread adoption. [5]

Fig. 2: Robot performing bin picking

Startups are playing a critical role in enhancing the utility of A.I. for practical industrial robotics applications. This is evident by the diversity of robotics startups that showcased their products and services at Automatica 2022, which is a trade fair held every year in Munich, Germany. Automatica is the leading trade fair for robotics and smart automation and features the world's largest range of assembly and service robots with more than 570 exhibitors from 35 countries and 28,000 visitors from around 75 countries. A compilation video from Learn Robotics and AI is embedded below. [6] [7]

Based on my visit to Automatica 2022, I shortlisted and created a list of the top 10 German robotics startups that work on the most cutting-edge AI technology.

I would like to point out that this list is meant to educate students who are pursuing (or aiming to pursue) studies in the field of robotics, computer vision, machine learning, or any other sub-field of artificial intelligence. Having an overview of the ongoing work in the industry is absolutely critical in not just finding a job but also narrowing down your own interest and choosing the right subjects and skillset to focus on. Needless to say, working at a startup is the quickest way to expand your skillset while making a meaningful contribution to the industry.

To get the most out of this article, follow the following sequence:

1. Read the full article
2. Go to the website Learn Robotics and AI and select the Industry anchor menu (Fig. 3)

Fig. 3: Select the Industry anchor menu on the website

1. Go to the company page of interest
2. Check out the primary products and technology stack of the company
3. Visit the Careers section on the official website of the company
4. Note down the skill set required for jobs of your interest

At this stage, you would be equipped with a broad overview of the ongoing work in the industry, the kind of products and technology you can expect to work on, and the kind of skillset that is required. It is now that you can finally create a customized project-based learning plan for yourself that has a high likelihood to land you a job in the industry of your choice.

Begin with the end in mind - Stephen Covey

Conventionally, the education system is structured to bombard you with the theory of all the topics, which is perhaps the right method if you need to have a research career. However, if your goal is to get a job in the industry as an engineer, more emphasis needs to be laid on the skillset that would be required to perform the job properly. This can only come by understanding the technology stack used in the industry and upgrading your skillset by taking up projects and learning only the necessary theoretical concepts along the way.

On the website (Learn Robotics and AI), you will be able to find the exact project-based goal-driven framework you need, to create your own personalized curriculum. Go to the website and click on Start here to check the framework. (Fig. 5) [8]

Fig. 5: Complete framework to start learning Robotics and AI

Now that you know how to use the list, let's proceed with the top 10 robotics startups in Germany that work on cutting-edge AI technology for robotics. Note that the following list only consists of startups located in Germany that work at the intersection of robotics and A.I., i.e. use machine learning, computer vision, planning, etc. for robotics applications. A more comprehensive list is available on the website.

Here is a short disclaimer before we proceed.

This list is created based on my own personal experience of either working at some of the companies, talking to people who work there, or just impressions I got from their official websites and/or online presence on social media and is in no way absolute. Additionally, the list only takes into account startups based in Germany that exhibited at Automatica 2022 and does not include other European startups which might be working on equally exciting technologies. The list is also not in any particular order and is not meant to be a comparison of startups with one another. Rather, this list is intended to be used by students to identify and define their sub-domain of interest in robotics.

Top 10 most cutting-edge German startups at Automatica 2022 that develop AI for robotics

• Truphysics GmbH

Fig. 6: Truphysics GmbH - Company profile

TruPhysics is a startup located in Stuttgart with approximately 15 - 20 employees and offers a variety of robotics products for intralogistics, manufacturing, quality inspection, and even households (Fig. 6).

Their primary products include TruShelf - which is a multi-level carrier transportation system, TruFleet for autonomous milkrun coordination, TruRunner - which is an AMR with lifter and rotation table, TruLife, ARTour Assist, Robert M1 and M2, and Kinova Kit.

The startup uses conventional computer vision technologies for marker positioning and object tracking as well as AI-based vision approaches for face recognition and object detection. They also use specialized range and tactile sensors in conjunction with the vision and AI systems for cases like person following and natural language processing for interaction with the human. [9]

• ArtiSense GmbH

Fig. 7: Artisense GmbH - Company profile

ArtiSense offers high-performance camera-based positioning solutions primarily for robotics and vehicles. Their main expertise is in visual localization and mapping and calls their technology ArtiSLAM which combines visual-inertial odometry, sensor fusion, and AI and is robust for outdoor dynamic environments.

It is based in Munich and has 20 - 25 employees. For more details about the company and about ArtiSLAM, you can visit their website. [10]

• Robominds GmbH

Fig. 8: Robominds GmbH - Company profile

This startup located in Munich, with the primary focus on industrial automation uses machine learning, computer vision, and robot control to allow anyone and everyone to program industrial robots and use them quickly and efficiently, particularly in a production or logistics setting.

Robominds have developed an intuitive interface for robot control that allows simultaneous control of several robots at one station. They use vision technology coupled with machine learning algorithms for grasping and bin picking operations. Robotbrain.fleet uses AI for route planning and intelligent fleet control for robots. [11]

• Sewts GmbH

Fig. 9: Sewts GmbH - Company profile

Located in Munich, this startup comprising 20 - 25 young people aims to solve the hardest challenges in robotics dealing with deformable object manipulation. They use computer vision and machine learning algorithms to automate the handling of deformable objects.

Their primary application is that of industrial laundry for which they offer solutions for handling towels from a bin to feed them one at a time to a folding machine. I actually did my Master's Thesis at this company and Sewts is one of the most pleasant places to work at. The co-founders of Sewts have talked in more detail about their technology and work culture in a podcast and a meetup session. [12]

• Gestalt Robotics GmbH

Fig. 10: Gestalt Robotics GmbH - Company profile

Located in the capital city of Berlin, Gestalt Robotics with 35 - 40 employees builds technology for intelligent industrial automation. They develop 2D and 3D image processing algorithms based on both conventional computer vision and machine learning. Their vision technology stack consists of image classification, detection, and semantic segmentation, pose estimation, activity recognition, and anomaly detection.

They also provide path planning, collision avoidance, advanced control, flexible grasping, SLAM, and AR/VR among other things. Their applications include quality inspection, visual asset detection, and inventory. [13]

• Micropsi Industries GmbH

Fig. 11: Micropsi Industries GmbH - Company profile

Another Berliner startup with 50 -60 employees, Micropsi Industries GmbH takes care of assembly, handling, machine tending, quality inspection, dispensing, and several such operations for manufacturing and intralogistics.

Micropsi uses technologies like machine learning, computer vision, and force perception to allow you to automate your workstation yourself quickly and inexpensively. Its product MIRAI, enables industrial robots to handle variance in production. MIRAI-powered robots use cameras and sensors to react in real-time to dynamic conditions in a workspace. [14]

• Mech-mind GmbH

Fig. 12: Mech-mind GmbH - Company profile

With 80 - 100 employees, Mech Mind delivers complete and reliable solutions for bin picking, order picking, depalletizing, palletizing, machine tending, assembly, gluing, defect detection, etc.

It offers an industrialized 3D camera, a machine vision GUI, a robot programming environment, and a deep learning platform software. Mech Mind primarily works with industrial robots and uses 3D vision, machine learning, and control-based approaches for intralogistics and automotive applications along with construction machinery and home appliances. [15]

• Neura Robotics GmbH

Fig. 13: Neurarobotics GmbH - Company profile

One of the most exciting startups with its highly stylized robot products Maira, Lara, MAV, MiPA, Neurarobotics builds a domain of cognitive robots for human-robot collaboration.

Its primary technologies include computer vision, machine learning, path planning, and obstacle avoidance. At Automatica 2022, Neurarobotics had the biggest and most grandiose of stations. With 80 - 100 employees and based in Metzingen, it is one of the most exciting startups to work at. [16]

• Magazino GmbH

Fig. 14: Magazino GmbH - Company profile

Located in Munich with 100-120 employees, Magazino GmbH is one of the top robotics startups in Germany.

It primarily develops mobile robots for manufacturing and intralogistics and uses machine learning, computer vision, and autonomous navigation along with advanced cooperative robot operating system. The main mobile robots developed by Magazino are Toru and Soto. For more information, refer to their official website. [17]

• Agile Robots AG

Fig. 15: Agile Robots AG - Company profile

Spun off from the Institute of Robotics and Mechatronics at DLR DLR's Robotics and based in Munich, this is one of the fastest growing startups which has recently achieved the status of a Unicorn, which implies that the company reached a billion dollar valuation. It currently has approximately 150 employees.

Agile Robots AG uses cutting-edge deep learning and computer vision technology for several applications. I personally did an internship at Agile Robots and based on my personal experience, it is definitely an excellent place to work if you want to work on cutting-edge AI technology for robot manipulation. [18]

This was the list of the top 10 German startups that exhibited at Automatica 2022 and use the most cutting-edge AI technology for robotics applications. For a more detailed list, visit the website. If you find this article valuable, I would appreciate it if you could offer your feedback in terms of likes and comments. (I will use the feedback to choose the topic for the next article). Also, feel free to request any specific topics you want me to cover through a video or an article. I aim to create a personalized all-encompassing system for anyone to start learning robotics, computer vision, and machine learning, and am open to covering any ideas or concepts related to the topic.

Author :

Chandandeep Singh

References :

1. Wikipedia.org - Unimate
2. Invent.org - The invention of the industrial robot
3. RobotsDoneRight.com - Most Popular Robot Applications in 2022
4. OnlineMasters.Ohio.Edu - 5 Industries utilizing robotics
5. RoboticsAndAutomationNews.com - How AI Affects the Robotics Industry and What the Future Holds
6. Messe-München.de - Automatica 2022
7. Automatica-Munich.com - Final Report
8. Exhibitors.Automatica-Munich.com - Exhibitors & Brands
9. Truphysics.com
10. Artisense.ai
11. Robominds.de
12. Sewts.com
13. Gestalt-robotics.com
14. Micropsi-industries.com
15. Mech-mind.com
16. Neura-robotics.com
17. Magazino.eu
18. Agile-robots.com

Image Sources :

Fig. 1.1: Wikipedia.org - Unimate

Fig. 1.2: ResearchGate.net - Implementation and Evaluation of multimodal input/output channels for task-based industrial robot programming

Fig. 2: Robotiq.com - How Robotic Bin Picking Affects Your Other Processes