Back in January, Hexagon, a global hardware and software reality technology company, announced the launch of Sixth Sense. Billed as a “disruptive open innovation” startup platform, the software was designed to rocket Industry 4.0 technologies into the future with the help of startups. The first innovation area challenge, artificial intelligence (AI) for sustainable smart manufacturing, recently wrapped up after an eight-week training program with two joint winners named: RIIICO and SmartParts.
Engineering.com recently spoke with Jan Büchsenschütz, RIIICO cofounder and chief marketing officer, to learn more about the company’s technology and how it aims to help change manufacturing.
“It’s a little bit like The Sims but on a factory level, moving around objects, replacing them, and getting a different expert to import their data into our tool,” Büchsenschütz said. “For us, everybody could just log in the browser and then just have the 3D model that doesn’t require any training of any sort. It’s supposed to be fun and awesome. There are so many challenges in manufacturing, which need to be addressed from sustainability and efficiency approaches. We can get it all together on one table and have fun doing it building an easy product. That’s what we do with RIIICO.”
One of the main challenges RIIICO’s founders sought to address was the need to change the factory floor quickly. From customers wanting their products fast to the life cycles of products constantly changing, the factory floor needs to be fluid. Traditionally, that has not been an easy option since the manufacturing space doesn’t constantly change with the products being made. In manufacturing, that may quickly lead to losing a competitive advantage.
RIIICO was designed to be a drag-and-drop software for a virtual factory floor. It uses AI to create true-to-life simulations from 3D scans. It enables collaboration among the factory teams and provides flexibility to enhance floor designs for improved, greener workflows.
“We generate these models using AI and the scanning data,” Büchsenschütz said. “We have a deep-learning algorithm that can detect objects based on the geometry of thousands of objects. Our AI looks around the texture and geometry of an asset and then next [makes a] prediction of how this error might be fixed. For example, there’s a hole in the floor because there was some machine scanning there, which we have moved away, so we look around the texture of the geometry. We make a prediction and close the floor. This way, you have reliable data you can trust and still work with.”
“There is no availability for a computer to understand that this is a robot, this is a cage, this is the ceiling,” Büchsenschütz said. “As a human, we understand this because we have a visual perception of it, but a computer only sees pixels and space. We go through millions of pixels really, really fast for deep learning AI. Based on their geometry, we make a prediction of what kind of object that will be.”
From there, RIIICO’s technology provides segmented objects on a list like an object tree of thousands of objects detected. Meshing is then used to generate surface models so that the model is no longer a cluster of points. Those models can be exported or imported into other software, making data that was once slow and difficult to work with smooth and available at any point of the workflow. While that might seem like the end, RIIICO also developed completion AI to account for any errors during the scanning process, such as a missed corner.
From the beginning, the RIIICO team set out to make something different than any 3D software available. The founders wanted their software to be available from any browser, workstation or laptop in any location. Another component was ensuring that it was easy and fun to use.
“If you think about 3D software, for some, you require a Ph.D. to be able to use this data,” Büchsenschütz said. “A lot of people are inhibited by this. They don’t have the time, or they don’t have the know-how to access this. With RIIICO, I always say the day we have to write a tutorial for our product, we’ve done something wrong. It’s supposed to be really smooth and easy. It’s available for everybody to work with these models—from the CEO to the worker in the factory. I think that is really something that sets us apart from anyone else in the 3D space.”
The opportunity to participate in the Sixth Sense platform has given RIIICO support and training that is invaluable. As a young startup, the team doesn’t have decades of manufacturing experience. Through Sixth Sense, they have been able to build relationships and learn from people who have that needed expertise. The feedback gained from participating in the platform has enabled RIIICO to make further enhancements and gain some ground in the industry.
“We are in a closed beta with customers like Hexagon and the validation of our product,” Büchsenschütz said. “Our beta software will be open by January next year, hopefully, to the public to be used. We’ve also been to the Berkeley SkyDeck accelerator in Silicon Valley because we said manufacturing is global. As a German startup, we needed to go to the U.S. to be where the innovation is happening. It’s been an amazing journey so far.”
RIIICO’s journey may just be starting, but the company is already focused on the future, which they say is all about collaboration and transparency.
“Manufacturing is so complex that we will need to collaborate with each other if we want to overcome all these challenges,” Büchsenschütz said. “Hexagon has awesome hardware. Hexagon has awesome software. I can now take the RIIICO model and directly pitch it to the simulation software and people can start working there. For RIIICO, this means we will be the experts generating these models out of raw scanning data and making it available to anyone who needs this data to work with—simulation people, logistics people.”
Sixth Sense is pushing forward and plans to have approximately two challenges per year. The company is taking proposals from startups for innovators in sustainability, machine learning, big data, sensors and robotics.