Konica Minolta develops AI-based solution "Suspicious behaviour detection" for its customer CZVV

| 12 October 2020

The solution is designed to be used with Mobotix smart cameras to prevent unauthorised access to confidential exam tests

Konica Minolta Global R&D has developed a new solution, "Suspicious Behaviour Detection", for its customer CERMAT – Centrum pro zjišt'ování výsledků vzdělávání (CZVV), based in the Czech Republic.The solution aims to help CZVVkeep graduation and entrance exams secure and will use a combination of Konica Minolta Mobotix smart cameras and artificial intelligence (AI) applications to do so. Specifically, the new solution is designed to prevent potential unauthorised access to confidential exam tests during printing and increase the level of security and overall efficiency of printing operations. The intelligent camera solution will automatically detect suspicious behaviour and trigger an alert. This eliminates the need for manual monitoring, is highly accurate and has a very low error rate.


The state examination institute CZVV is responsible for preparing the annual upper secondary school leaving exams (Maturita) and publishing exam results for all schools in the Czech Republic. Each year, CZVV prints hundreds of thousands of assignments for secondary-school graduation in a very short time, which are then distributed to schools.
 

Printing the tests on the Konica Minolta AccurioPress 6136 monochrome production printing system is a very sensitive process that is subject to the strictest security requirements.
 

Such an approach was required to prevent security breaches such as secretly photographed the test documents or spending too long in the vicinity of the printing machines and was a highly elaborate process.

 

Konica Minolta worked together with CZVV to precisely define a variety of situations for which a notification should be sent in advance of rolling out the new system.
 

The major advantages of the AI solution are the high level of precision and significantly reduced susceptibility to errors compared to manual monitoring by an employee. 

 

"Our application is based on the principles of AI technology, computer vision and machine learning, and can enhance or replace the work of the operator to make security operations more efficient and save on costs. The entire solution is fully compliant with GDPR", acknowledges Tomáš Slavíček.
 

Konica Minolta systems do not just cover what is going on around the printers — they also monitor the status of the printers themselves. If the warning light for the printing system is red (e.g. printing is blocked due to an error) or yellow (e.g. an error has occurred during printing) rather than green (everything is running smoothly), the AI solution and Mobotix camera can trigger an alarm so that the service technician can correct the problem.
 

Konica Minolta Global R&D has also worked on another AI application that works with Mobotix thermal imaging cameras: As part of its Infection Management portfolio, the digital workplace provider has developed the "Temperature Screening App" to detect increased body temperature. When individuals pass through the camera's field of view in offices or other facilities, their thermal images are analysed and any elevated body temperatures exceeding a specific threshold are identified. Those concerned will be warned and the relevant personnel will be notified so that immediate action can be taken to contain the potential spread of the infection. The application has been available since October.
 

For more information about video solution services, please click here and for suspicious behaviour detection here.


quotation marks

We previously have provided this protection with a standard camera system that recorded the image but was unable itself to identify suspicious behaviour and raise an alarm. So the security control relied on the focus and attention of our employees, who, in case of anything suspicious, had to go through hours of recording.

Jindřich Mach

Director of the Internal and Manufacturing Infrastructure Section CZVV Czech Republik