Process monitoring

Continuous process with multi-dimensional time-series data (e.g. IoT data) poses a serious challenge for human interpretation. Often multiple indicators values from the process can be properly understood only by an expert who has been on the job for a long time. That job is also tedious and tiring.

With Ai4Value’s Process monitor the task is made easy. The flow of the process can be understood “at one glance”, and with short training. The Process monitor has following features:

• A map-like interface, where the processes’ signal time-series data space is compressed into 2D view with color coding different states of the process.
• The progress of the process can be examined and animated on the map for any desired time scale.
• Accepted limits for the process can be determined on the map, and deviations from those can easily be seen as animated states appearing outside the safe boundaries.
• Accuracy and disturbances to repeating sub processes are shown in color coded timelines.
• This allows for early detection of possible problems, such as maintenance need.
• It can also be used for performance comparisons between workers.

An example of the Process monitor usage is an analysis of three months of vibration data from an industrial CNC machine. The data consisted of ca. one million vibration measurements with frequency values on six channels from an IoT device.