With a huge amount of data, it is critical for humans to identify changes in performance and take actions on that insight.

Moreover, it’s hard to interpret the meaning of a specific shift in a metric, that could be innocuous or a life-changing opportunity for growth.

Veos Digital Anomaly Detection Service has been designed for those companies that want:

  • to discern between insignificant events and those that are truly unusual, calling for actions
  • to let people be focused on strategy, reducing time-consuming activities
  • to have a unique interpretation of data shared between teams

Anomaly definition

Anomalies can be classified into two categories: 

  • Expected Anomalies: they are generally well understood and easily connected to specific events
  • Unknown Anomalies: they are not understood and humans cannot predict them

Veos Digital Anomaly Detection Service is referring to Unknown Anomalies.

Characteristics

It identifies unexpected data points deviating significantly from expected behavior. Only anomalies are shown to users – reducing the cognitive load and allowing them to tell the signal from the noise.

It can be applied to data with high quality and one of the measurements within the data contains sufficient information to reveal the anomaly, such that a human operator, given sufficient time and skill, would be able to unearth it.

It can be also applied on data set whit a new key variable to the mix – time.  In many applications, they are vital for representing the normal behavior of a system.

Scientific base

Veos Digital Anomaly Detection Service is based on Time Series Analysis.

It exploit Unsupervised learning technique, so the system intelligently assess the datapoint that are likely to be historical anomalies, and return that to the user, alongside determining this for future datapoint.

Use cases

Anomaly detection can be used also for marble issues identification.

Here’s an example of few sample of marble photos analyzed to find the imperfection.

Marble anomaly detection

On the other side, we can use AI to collect anomalies in Agritech industries. In this example, the algorithm is highlighting the leaf rice with a potential HISPA (Dicladispa armigera).

Rice HISPA anomaly detection

Integration

Our system is RESTful APIs based. It’s a standard adopted by many Software producers to provide applications to Third Parties.

Simple integration

Zero maintenance

Pay per use