Changing the way industry thinks about anomaly detection

Our trusted technology is transforming the way industry handles anomaly detection. Automated anomaly detection can drive efficiency gains and reduce the likelihood of error.

  • Automotive

    We detect faulty components for car manufacturers before they make it into final production vehicles.

  • Manufacturing

    We assess real-time production data to detect inconsistencies between manufactured parts.

  • Health technology

    We detect irregular diagnostic device measurements resulting in higher quality patient care.

  • Connected devices

    We can analyse connected device data in real-time. This enables us to detect compromised devices.

Automotive R&D Testing

Industry need

As part of the development process for automotive heaters, each heater is put through extensive test cycles to ensure the highest quality standards. Endurance tests run for over 2400 hours and involve 150 different signal inputs. This makes data analysis very difficult.

Applying can analyse all data generated during the testing process to identify patterns formed by correlated data. This allows engineers to focus their attention on the problematic sensor channels and complete an expedited root cause analysis. Once engineers are notified about an anomaly they are able to address the issue before continuing the heater test, saving time and resources.

Valuable outcomes demonstrated significant improvements over the previous rule-based alert system:

  • 85% reduction in time spent on daily data inspection by engineers
  • Analysis of all data versus previous method which analysed 10% of data
  • Faster root cause analysis
  • Identification of 5 previously unknown anomalies across 9 devices
  • Anomalies were found in sensors that were previously only observed manually

Manufacturing quality assurance

Industry need

It is challenging to determine if a plastic part has been correctly made until after the production process. This results in a loss of time and resources as faulty production is not halted and additional components must be recreated after the fault has been detected.


Using pressure data generated from sensors on plastic molds, can analyse multiple channels to quickly identify if a plastic part is not forming correctly. This could be due to an air bubble or a misshapen part. After triggers anomalies, engineers can stop the process and ensure that the part is not sent off to the next painting stage, avoiding a further waste of resources in the production line.

Valuable outcomes

Using, manufacturing companies have real-time insight into the precision of the production process. Once anomalies are identified, engineers can tackle the problem immediately, ultimately saving money, time and resources.

Diagnostic machine maintenance

Industry need

Traditionally, a diagnostic machine manufacturer will service a machine on a scheduled basis. Data from the machines is sent to the cloud, but no real-time analysis of this data is performed.

Applying can analyse data generated by the machine and perform multiple channel analysis to identify anomalies that indicate reduced machine functionality. Based on the identified anomalies, the lab will be notified about the potential for irregular measurements produced by the machine and can take action by re-running measurements, calibrating the machine or informing the technician that a repair is required.

Valuable outcomes

Anomaly detection within predictive maintenance of a diagnostic machine can lead to the following outcomes:

  • Reduction of service costs by prioritising maintenance
  • Reduction of costs associated to poor quality test results
  • Efficient planning of technician maintenance schedules

Trusted to deliver

Ready to get started?
Get in touch for a demo
Get in touch
Send a message Paul-Lincke Ufer 8B, 10999, Berlin 🇩🇪

© All rights reserved.