The emerging startup, founded by one of OpenAI's first engineering hires, aims to “redefine manufacturing” with an AI-powered factory to create bespoke precision parts.
The company, named Daedalus, is based in the southwestern German city of Karlsruhe, where it currently has a standalone factory. Here, Daedalus receives orders from industries such as medical devices, aerospace, defense, and semiconductors, each of which requires unique components for their products. For example, a pharmaceutical company may require a customized metal casing for a valve used in the manufacture of a particular drug.
Daedalus, which is looking to ramp up operations with a view to opening additional factories in the domestic market, today announced a Series A round led by Nokia-backed NGP Capital, alongside existing investors Khosla Ventures and Khosla. Ventures, added.
This brings Daedalus' total funding to over $40 million, with the addition of other prominent investors including Y Combinator (YC), which became involved after Daedalus participated in YC's Winter 2020 program.
Daedalus factory floor image credits: Daedalus
fragmented manufacturing
The manufacturing industry, especially those related to the production of precision parts, is by all estimates highly fragmented. While it is tempting to imagine that a typical manufacturing setup in 2024 will resemble a large-scale automotive assembly plant, this is actually only true if it involves mass-produced products (such as cars). In reality, it's actually a little different. Precision parts used in industrial machinery.
A company that has been designing industry-specific valves for decades probably won't manufacture them all in-house. They typically rely on a traditional network of manufacturers. This may mean working with a small business consisting of one professional “artisan” and a few helpers working in a small facility.
“What this means is that they are not doing much in terms of digitalization. Basically they are just used to working with paper and pen, so it is difficult to change that.” Daedalus Jonas Schneider, founder and CEO of , told TechCrunch. “This means that very low-tech manufacturers are supplying the most critical components to these very high-end products.”
Daedalus Founder and CEO Jonas Schneider image credits: Daedalus
Founded in 2019, Daedalus uses similar off-the-shelf hardware available from any manufacturer, but its special features are deployed at the top level to control and optimize the 'shop floor'. It is located in the software. This means automating many manual tasks. Involved in manufacturing specific parts. So customers submit their CAD (computer-aided design) drawings as usual, and Daedalus injects automation into the process to develop these drawings into finished parts.
“It's coordinating all the workflows, from production to planning to scheduling people running around the factory floor,” Schneider said.
For context, when a machine starts producing a new “part”, what tooling is typically needed and what settings are used to create the exact shape and dimensions of the part? It involves dozens of steps and hundreds of decisions that affect the and so on. And here he enters Daedalus. The company's software captures manufacturing decision data for one “part” and uses it to guide decisions about how similar parts are made in the future. Therefore, a slightly larger valve, or a valve with additional fittings, may be virtually the same as the previous part, so Daedalus uses pattern matching to apply that previous knowledge. , configure the machine to suit the new parts.
Daedalus extends the basic concepts of 3D printing, which has democratized manufacturing processes for more than a decade, in many ways. But with machine learning smarts built inside, things are being taken to the next level. It's like 3D printing him on steroids.
“This comparison is very apt. For me, as an outsider to the industry at first, custom manufacturing [already] The solution was 3D printing. But it mainly comes down to the technical limitations of the process,” Schneider said. “Even with 3D printing, new parts have to be specifically designed to be 3D printed, which is actually a fairly expensive process. But for a large part of the industrial base, that's not really feasible. 3D printing is not possible because it is not precise enough or the material is not strong enough.
In some ways, you can think of what we're doing as taking this idea from 3D printing and applying it to high-end, industrial-grade parts. ”
The story so far
Prior to joining Daedalus, Mr. Schneider served as Head of Technology at OpenAI, where he helped launch the company's robotics division in 2016. In fact, while OpenAI may now be best known for its flagship product, his ChatGPT AI chatbot, the company also operated a robotics division. A project that conducted research such as solving a Rubik's cube using a robot hand. Schneider was directly involved.
OpenAI Rubik's Cube Hand image credits: Open AI
Although OpenAI ultimately disbanded the team in 2021, Schneider spearheaded the software engineering side of operations for more than three years before leaving to launch Daedalus in 2019.
Although there were various reasons, Schneider eventually left to found his own startup, but an experience he encountered while building a Rubik's Cube hand played a part in his decision to launch Daedalus.
“At one point the robot hand broke down and we had to get spare parts,” Schneider said. “And guess what? It had to be precision manufactured. We had machines similar to the ones we use today, but we had to wait months to get these parts. And I thought, why is it so hard to get spare parts here? All of this led me to look a little bit more closely at this whole manufacturing space.”
Daedalus currently has a single 50,000 square foot factory in Karlsruhe, from which it primarily targets German-speaking markets, including Austria and Switzerland. In the short term, we plan to expand to a second factory in Germany and further afield if there is sufficient demand.
“This is a blueprint factory, right? Here we learn all the systems and all the knowledge and distill it into how to make these parts,” Schneider said. “And in the long run, we will be putting these factories where our customers need them.”