Manufacturing today requires more customization than ever, but traditional robots are only designed for repetitive tasks. For small production runs, programming a robot for each new part is often too time-consuming and expensive, leaving manufacturers with a difficult choice.

At the same time, European manufacturers face higher labor costs, worker shortages, and growing competition from Asia. For Daniel Seebauer, Co-founder of Batch Robotics, the answer lies in automation.  

After co-founding drag&bot in 2017 to make industrial robots easier to program, Daniel is now going even further with his new venture, Batch Robotics, by removing the need to program each component. The company uses 3D vision, simulation, and process-specific AI to build robots that adapt to each workpiece, making batch-size-one automation possible.

Read on to discover why Daniel believes AI could transform industrial robotics, what’s still holding physical AI back, and why automation could be key to keeping manufacturing in Europe.

About Batch Robotics

Batch Robotics develops AI-powered robotic systems for high-mix, low-volume manufacturing. Their technology automates surface-treatment tasks like grinding, polishing, sanding, and welding without having to program the robot for each part. By combining 3D vision, physics simulation, and process-specific AI with standard robotic hardware, Batch Robotics aims to offer systems manufacturers can use as ready-made products instead of custom projects. The long-term goal is straightforward: robots that can be set up on the factory floor and work like skilled employees, adapting to each task on their own.

From making programming easier to removing it altogether

When Daniel and his co-founders launched drag&bot in 2017, programming an industrial robot typically meant working with the specific language developed by its manufacturer. drag&bot tackled this by building a no-code platform that let factory workers set up applications, teach robots, and make adjustments themselves.

They helped make smaller batch sizes more economical to automate and anticipated a wider industry shift toward more user-friendly robot programming.

But Daniel noticed a key problem: even if the programming was easier, someone still needed to do it manually. For factories making only a few parts, spending hours setting up a robot for one small task isn’t worth it.

Batch Robotics takes a different approach. Rather than just making programming simpler, the team wants robots to understand tasks without any component-specific programming, giving customers a complete system that can perform manufacturing processes on its own.

“With Batch Robotics, we are now leveraging the technological capabilities that AI brings to the field of robotics and focusing on batch-size-1 automation, where zero robot programming for a specific component is necessary.”

Why small-batch automation is so difficult

Factories can be unpredictable places. Parts can be made in different places, have varying orientations, or have unique surface shapes. CAD files might also be outdated or missing entirely. Meanwhile, tasks like grinding, polishing, and welding involve physical forces that are hard to model. Traditional automation gets around this by designing robots for specific, repeatable tasks, which works well when it comes to producing the same product in large numbers.

Small batches change the economics because each variation requires more setup work.

AI is starting to address this. Modern GPU-powered physics engines let Batch Robotics simulate physical processes much faster than real time. Meanwhile, 3D vision and process-specific AI help robots recognize new shapes and plan their movements. The main challenge now is making these smart systems reliable for daily factory use. 

“The only thing that I believe is still missing is these new software stacks... making these systems intelligent, really reliable for factories.”

Factories don’t want another piece of software

One of Daniel’s main takeaways from drag&bot was that manufacturers don’t just want to buy software; they want to buy solutions to their problems. Many already struggle to find reliable workers and automation experts, so giving them another tool to set up only adds to their workload.

Batch Robotics now focuses on offering a complete robotic system for a specific manual task, instead of just a software layer customers have to integrate themselves.

This philosophy also underpins the technology that they are producing. Since internet connections in factories are not always reliable, systems need to keep working using local edge computing. They also need to be standardized.

Traditional automation often depends on system integrators designing custom solutions for each and every use case. Daniel believes robotics will only scale when systems are delivered like industrial products: you install them, plug them in, and let them do the job.

“You go there, you plug it in, and then it does the job like an expert worker. This is the way we imagine it.”

The economics of automation are changing

For manufacturers, automation is mainly about economics. Batch Robotics works with customers to see how its systems might impact labor, output, quality, and material use. Daniel says the company aims to deliver immediate savings of around 30%, depending on the part and the process, compared to current methods.

But cost is just one factor. Surface-treatment jobs are often tough, repetitive, and hard to fill. Daniel thinks this shifts the discussion about robots replacing workers, since factories already struggle to find people for these roles.

He expects workers will move toward running, maintaining, and improving robotic systems. Just as software developers now guide and supervise AI instead of writing every line of code, manufacturing employees could start managing automation at a higher level.

“No one in the future will want to do that job, will stand there at two AM in the morning and do a very dusty, dull, dangerous, hazardous job.”

Physical AI is impressive, but the demos are the easy part

Physical AI is progressing quickly. Vision-language-action models show robots learning tasks from people. Humanoid robots are also improving all the time. Still, Daniel is cautious about the difference between a great demo and a system that works reliably every day in a factory.

He calls this the 80/20 rule. Getting a robot to do the easy 80% of a task is often simple, but the real challenge is handling all the unusual cases that make up the other 20%. Industrial systems must work reliably when objects are in new positions, something is in the way, or unexpected things happen.

Daniel compares today’s excitement to the buzz around self-driving cars ten years ago. The technology is moving forward, but it has taken much longer to roll out than many expected. He thinks physical AI will develop in a similar way.

“There's a huge hype about what's now going on, and it's technically feasible, but it will take another decade before we see it actually really reliable on the shop floors, in our houses, in our factories.”

Robot hardware is becoming a hot commodity

As robot hardware gets cheaper and more standardized, Daniel expects software to become the main source of competitive advantage. Cobots are already becoming common, and humanoid robots could become another option once affordable and reliable. However, he believes the intelligence behind the robot is more important than its physical shape.

The bigger change will be moving from teaching robots for each batch to plug-and-produce systems. Instead of engineers programming every new part, AI-powered automation could scan the object, figure out its shape, and decide how to do the job.

Over time, Daniel expects physical AI to handle more complex tasks like assembly and welding where understanding physical limits is crucial.

“In the end, for me, a humanoid is just another kinematic. It's just another piece of hardware.”

Building for the factories that already exist

People often picture the factory of the future as a spotless, fully automated space built for AI and robots. Daniel thinks the reality will be much messier. Industrial equipment can last for decades, so manufacturers cannot just build new factories from scratch.

That’s why Batch Robotics is designing its systems for brownfield environments so they can work with existing machines, production lines, and processes. This also explains why Daniel thinks the change will take time.

As labor gets harder to find and global competition increases, the push to automate is growing. For tasks like surface treatment, he believes automation will be necessary to keep production affordable in Europe.

“Most of these manufacturing lines and the machines there are often there for decades.”

Can automation keep manufacturing in Europe?

This is where Batch Robotics’ mission extends beyond just robotics. European manufacturers face higher costs and strong competition from Asia, while global tensions and weak supply chains have made local production more appealing. Daniel believes automation can help Europe compete in a new way, not by making huge numbers of identical products, but by focusing on customization.

He envisions factories making thousands of different products cost-effectively, eventually reaching true batch-size-one production where nearly every item can be customized.

AI-powered robotics could give this high-mix, low-volume production the same cost advantages as mass manufacturing, making local production more competitive. Without more automation, Daniel thinks many processes Batch Robotics focuses on will move to cheaper markets.

“If we don't automate it, this will move away to probably Eastern Europe or even Asia. There's no way we can keep up in terms of production costs and quality.”

From mass production to mass customization

Daniel’s long-term vision is for manufacturing to become more local, flexible, and automated. Robots take on more dangerous, dirty, and repetitive physical work, while people focus on maintaining systems, optimizing processes, and introducing new workflows. 

Most importantly, customization no longer has to come at the expense of efficiency. If robots can adapt autonomously to different components, producing five variants could become almost as straightforward as producing 5,000 identical units.

Batch Robotics wants to build the intelligence layer behind that transition, transforming standardized hardware into autonomous, zero-programming systems. Its initial focus is surface-treatment applications such as coating, grinding, deburring, and welding. Still, the ambition is much bigger: to eventually power more than half of the world’s surface-treatment automation.

“All the manufacturers here in Europe will produce batch size one for their products as well.”

Quick-fire round

A tool you can’t live without?

“Claude and Gemini. These are tools that we use extensively all day, every day.” 

How do you recharge?

“I go for a run with my colleagues in the morning every week. Spending time with my family and kids is also very grounding for me when it comes to work and also the aspirations that I have as a founder.”

Founder or start-up that you admire:

“I really think Sereact is a company to watch out for. I think they have a similar approach, focusing on software, building the intelligence layer for intralogistics. Being much more KPI-driven and focused on what they achieve and customer-focused rather than blah blah big vision.”

Advice that stuck:

“In the end, everything, even though we're talking so much about technology and how AI and everything changes, it's still all a bit of people's business. It's about building trust, building relationships, giving back to people, then it somehow naturally comes back to you in another way.”

Making automation the standard

Daniel’s vision for robotics is to make automation feel less like a complex engineering project and more like a standard product.

Batch Robotics wants manufacturers to install a system, plug it in, and let the software handle the rest, adapting to different parts without needing to program each one. 

That future will take time to arrive. But as skilled labor becomes harder to find, customization grows, and European manufacturers face more global competition, the need for flexible automation is already clear.

Batch Robotics believes that someday, programming a robot for each new part will seem as old-fashioned as programming any modern machine by hand.

Curious about how Batch Robotics is making batch-size-one automation a reality? Visit the Batch Robotics website to find out more.