Powerful camera-based AI applications

Joining forces in enterprise resource planning and automated image processing to create powerful AI solutions for industry, logistics, and retail. 

The partnership between Rutronik and collective mind already has an impressive track record, with initial projects in component logistics and a focus on the traceability of more than 100,000 electronic components.


    "We are joining forces with the aim of offering internationally scalable software and
    hardware solutions developed from a single source for AI applications in retail and 
    industry."

    Artur Hefner, Managing Director at collective mind

Electronic component logistics use case with > 100,000 parts

Logistics experts assume that much of the scanning technology currently in use will be supplemented or even replaced by camera technologies in the future. At Rutronik's largest logistics center, AI-supported image processing is already being tested in logistics in order to accelerate the growing demand for traceability at the product level. 

Rutronik, with its more than 100,000 electronic components and the associated variable recording data, is an ideal use case for testing this development. The goal of both companies is to use the AI application in logistics in the future fully automatically and with flexible, reliable information recognition, even with changing formats and arrangements of product data on the corresponding product labels. The first pilot of the development is already running at Rutronik's largest logistics center in Eisingen near Pforzheim.

The initial project successes achieved by Rutronik and collective mind speak for promising AI-based logistics solutions that, together with other AI innovations, are intended to drive AI penetration, particularly in German SMEs.

  • Complexity in goods receipt due to a wide variety of products and formats as well as documentation standards cause increased manual activity
  • Visual impairment to the point of partial illegibility of documents, e.g. QR codes, make automatic capture difficult
  • 24 percent of QR codes are unreadable
  • Changing lighting conditions from very strong to very weak with interfering reflections are not ideal for scanning projects

  • No need to customize environments and workstations to achieve automated capturing
  • Optimize QR code reading rate to 99.8 percent
  • Capture and validate bills of lading up to 10 times faster and with more information
  • Automatic reconciliation of data management details (ERP interface)

  • Speed up processes, reduce manual steps and minimize susceptibility to error
  • Optimize QR code reading rate (standard approx. 75 percent)
  • Read multiple codes simultaneously

The AI solution uses real-time images from an industrial camera to recognize and count goods or delivery notes, and to read, complete, and store all the information from the labels.

Hardware meets software expertise for compute-intensive AI applications

In addition to its logistics and consulting expertise, Rutronik brings hardware expertise at the product level of electronic components to the partnership. This will be combined with collective mind's software expertise in AI-based image processing approaches. 

Specifically, collective mind aims to become the international number one for these AI solutions in the field of machine vision applications in industry and trade, with the expertise of Rutronik as a long-standing distribution partner of many medium-sized companies in Europe, Asia, and North America. With development projects in the industrial environment, especially in the field of vision robotics, the AI specialist is pushing ahead with great growth potential.

  • Manual sample-based visual inspection is inadequate and time consuming
  • Competitive QM products require input from historically collected defect types
  • Extremely high number of different, complex, known or unknown defect types or even unknown defect types makes it difficult to predefine and up-to-date rules and reference values for anomaly detection
  • Analysis to draw conclusions about the causes of defects
  • Traditional systems typically only check individual parameters (e.g. temperature, pressure, dimensions) separately

  • Automated, AI camera-assisted testing is more efficient and more robust
  • AI solution learns with normal and good parts, no input from already existing or known defect types or anomaly parts necessary
  • Automated conclusions about the worn tool involved in the machining process („predictive maintenance“)
  • Combination of various measurement parameters and sensor data (radar, LiDAR, ultrasound) simultaneously possible („sensor fusion“)

  • Minimize manual visual inspection effort
  • Eliminate the need for spot checks
  • Faster, more efficient conclusions about the causes of defects

Quality management currently works with defect types. AI-based anomaly detection enables inspection based on normal and good parts.



   "The trend is towards complete embedded solutions with powerful, compact hardware for 
   compute-intensive AI applications."

    Fabian Plentz, Chief Operating Officer (COO) at Rutronik


Interested in more insights?

Then join us at our joint event "AI Vision" in fall/winter 2024 and discuss your use case with us.

More details about the event will follow soon. You can already register here for the free invitation mailing.

Downloads

News "AI at Rutronik"

Corporate News on partnership with collective mind

AI Network Meeting at Rutronik


Get in contact with us

Micropages