ZEISS Robotic Start-up Challenge

30.06.2017

Win ZEISS as a capital investor

Participate in the Robotic Start-up Challenge

Submission of proposals by 31.07.2017
Pitch & award event in Berlin on 10.08.2017

  • Win ZEISS as a venture capital investor for your startup or to become a strategic partner for ZEISS.
  • Get a contract of 25.000 € to provide a prototype or technical contribution.
  • Become a trusted collaborator of ZEISS Corporate Research and Technology Network.
  • The possibility of getting hardware from ZEISS for the duration of the project.

Approximately 10 years ago Steven Phillips from NIST predicted that growing CPU performance and resolution of digital cameras will allow to measure objects without any other tools by simply walking around and taking pictures. Up to now, such solutions are not robust and precise enough to cover the needs of our customers in industry (for instance car body manufacturing, aerospace, power plants…). On the other hand, such solutions could create a low end disruption of the existing methods based on precise coordinate measuring machines, Laser Tracker based solutions.

Since the core of measuring technology moves towards software for evaluation of measured data, such methods gain growing interest. Further advantages of such methods are the flexibility since in most cases objects are heavy and difficult to transport.

A cheap camera based solution could potentially become an attractive tool in smaller workshops for instance for repairing cars. As soon as metrology becomes cheaper and more flexible we see large new markets, which are not yet using our products.

Your challenge:
Camera-based 3D Metrology of large objects
We ask for a camera based solution for the 3D-digitalization of industrial objects and potentially complete factories. The method should therefore be scalable. The proposed method should be based on camera images without exact knowledge of the camera position and should work handheld. Any kind of camera technology may be used (including RGB, stereo, ToF, LIDAR etc.); number of images and cameras is not limited but measuring time (including evaluation) should not exceed 10 minutes. Illumination, wave length, type and number of cameras and size are not restricted as long as the cost limit (see below) and a mass of 2 kg are not exceeded. IMU use would be acceptable. The overall system concept outside the scope of the challenge plans to integrate the camera-based 3D digitalization system on a mobile robot in an industrial environment.

The result should be a closed 3D mesh of the object.

Accuracy
As good as you get towards submillimeter range; goal: Better than 0.1 mm (3 sigma)

Measurement volume: 10m³

Measurement object
Steel storage rack. The object to be measured needs to cover a reasonable volume. For ground truth generation, it must have features which can be measured up to the claimed accuracy by other methods (e.g. CMM, pattern projection, Laser). The object should have glossy and scattering surfaces. If necessary, Zeiss can provide support for measuring the object by means of a coordinate measuring machine.
Estimated cost of components (except for computer/GPU... and robotic mover (if applicable)) smaller than 2500 € (evaluated at 100 pieces annually).

Key performance indicators
1. Accuracy
2. Cost
3. Measurement time/duration

PLEASE CLICK HERE FOR MORE INFORMATION

Got More Questions?

Contractual requests
Dr. Danny Krautz
danny.krautz@zeiss.com

Technical questions
Dr.-Ing. Frank Höller
frank.hoeller@zeiss.com

Submission
Dr.-Ing. Frank Höller
frank.hoeller@zeiss.com