subscribe

Subscribe to ZEISS Stories:

Get the latest ZEISS Stories straight to your inbox.

Robots Given the Gift of Sight

From measuring a forest to self-driving cars, start-ups can achieve a whole lot thanks to 3D mapping. In Berlin, five start-ups presented their ideas for a cost-effective and precise measuring device.

A risky prediction was made by a man whose work is centered on precision. Roughly ten years ago, Steven Phillips suggested that it would soon be possible to measure large objects in 3D – simply by walking around them and taking photos. Phillips works for one of the most precise authorities in the world, the National Institute for Standards and Technology in the USA. Ten years ago, a scenario in which users simply take photos of their car to generate highly precise 3D images seemed as likely as getting drones to take your shopping home.

At the time, hardly any of Phillips’s colleagues even dreamt that we would now be buying smartphones, which are capable of visualizing their surroundings in 3D. The accuracy of procedures like these does leave something to be desired:

Dr. Frank Höller, Principle Scientist at ZEISS

“The apps are good fun. But I wouldn’t use them to install a kitchen and certainly not to measure cars or planes.”

In the industry, the 3D scanners can take measurements that are accurate down to the micrometer – but most of them are either tied to a particular location or are too bulky to carry around a car. And devices that can accurately measure large objects and are portable do cost a pretty penny.

Challenge aims to bridge market gap

“What we’re looking for is a device that meets our three criteria, i.e. is highly precise, cost-effective and portable,” says Dr. Kai Daniel, Director of Systems and Engineering at ZEISS. To bridge this gap in the market, ZEISS has come up with the Robotic Start Up-Challenge. The task? To design a device that costs less than €2,500 to produce – and can digitalize a warehouse rack in 3D, and measure it, in just ten minutes to an accuracy of 0.1 millimeters. Unlike the majority of high-precision 3D scanners, this solution is not intended to work with reference markers, but be camera-based. So Steven Phillips’s vision has in fact come to life: to simply carry a camera system around a room – no preparation or adjustment necessary.

A host of applications for 3D measurement

The start-ups are competing for the chance to win financial support and gain access to ZEISS’s hardware and professional network. The five teams that made it through to the final presented their designs on 10 August in Berlin before a panel of ZEISS experts. All teams worked with 3D measurement but focused on different areas: start-up Vins from Berlin has developed a process for taking 3D photos of forests and using this data to determine the position of the trees and the thickness of their trunks. Iviso from Vienna is working on software that helps self-driving cars find their way. Gestalt Robotics is developing processes that enable robots to recognize their surroundings.

The teams pursued different approaches in a bid to win. Gestalt Robotics presented a process that takes photos of objects with a stereo camera and calculates the result using a high-resolution camera. HD Vision Systems from Heidelberg pursued a light field-based approach. “Not only are we capturing colors, but also the direction the light is coming from,” explains start-up founder Dr. Christoph Garbe. “This means we can also measure shiny objects with a high degree of precision.” The prototype for his product, Lumiscan+, is a device the size of a tablet, roughly 3 cm thick and features 13 built-in industrial cameras. “It will mean that robots will be able to see more clearly than people,” says Garbe. “And it will also see robots performing tasks they were previously incapable of completing. They will be able to detect and measure objects simultaneously.”

The winners

Not only did the ZEISS experts assess the teams’ methodical skills, they also had to decide whether there were enough different talents on the team to bring a product to market. The winners were announced once the presentations had been given and meetings conducted with each team. Gestalt Robotics came in third, and Iviso notched up second place. The overall winner was HD Vision Systems. “They had the most interesting concept and the skills to bring the product to life,” says Dr. Frank Höller. “I think this device has the potential to succeed.”

The winning team wasn’t just in it for the prize money, but also to benefit from ZEISS’s expertise and professional network.

Dr. Wolfgang Mischler, HD Vision Systems

“ZEISS is looking for a young and flexible team. We are looking for a partner with a good reputation to help us bring our sensor to market. I think we’ll make a great team.”

Further Information

(No Ratings Yet)
Loading...

Vote:
Shall we produce more on this topic?

20 0

Related Posts

2 comments

Prof. Kathleen Richardson

Very interesting? What about a challenge in the infrared – good metrology tools typically do not really ‘see’ what we need to beyond the SWIR – what about lower cost optics which offer improved SWAP from novel materials?

Is the challenge carried out annually? Will there be another one? Thank you!

Dear Prof. Richardson,

thank you for your interesting suggestions regarding the application of IR optics. We are happy you enjoyed the story.
This was our first scientific challenge and we are currently considering more. At the moment we do not plan a challenge in IR optics.
If you have a concrete application or technical solution (to realize cheap IR optics) we are curious to know.

Your ZEISS Stories Team

Leave a reply