Personnel Movement Tracking

Problem Statement

Monitoring of personnel movement in a zone, count of zone violations, objects identification in zones.

Solution and Benefits

VAST surveillance solution allows to easily define 2D zones or do-not-cross lines.

Continuous monitoring of violations is done by the platform without any user intervention

It can also identify abandoned objects

Attendance through Face-ID

Problem Statement

Face identification for attendance for multiple people at one go.

Solution and Benefits

VAST face-ID solution would be efficient.

Pollution Monitoring

Problem Statement

Monitoring of smoke coming out from chimney in a manufacturing unit.

Solution and Benefits

VAST was used to build a custom solution for a manufacturing unit to identify the smoke as and when it come out of chimney.

Alerts can be generated so that it can be managed and contained.

Security Management

Problem Statement

Monitoring of security guard presence in security room.

Solution and Benefits

VAST was used to build a custom solution for an IT park to detect the presence of security guard at their specified location all the times. Guard was detected through multiple features extractions.

An alert was generated if guard is not present more than a specified threshold time limit.

Classroom Monitoring

Problem Statement

Monitoring of classroom activity and teacher presence*

Solution and Benefits

We are developing VAST for monitoring classrooms for student activity and orderliness, and whether a teacher is present or not. This will help school administrators manage the timetable of the school properly.

Closing of the loop by alerting the person(s) incharge and subsequent verification that the alert has been worked upon.

Construction Site Monitoring

Problem Statement

Monitoring of personnel in the construction site for safety*

Solution and Benefits

We can work on monitoring the personnel in a construction site. We can train the neural network to track jackets, helmets and ensure the person is wearing the same.

*This use-case is under active development with a safety and security company.

Abnormal cells Identification

Experiment Description

A neural network was developed to count RBCs in human blood for a customer in diagnostic industry.

The same NN model is ported to identify nuclei in human cells as case study to demonstrate our capability in microbiology applications as real customer data can’t be shared.

With minor modifications in the NN models, it can be ported to other microscopic organisms for information extractions.