Logo-Pinnacle

MicroEDS

The “MicroEDS” is an Event Detection System (EDS), design to run on top of SBC (Single Board Computer). It is a software platform which utilizes a machine learning (ML) technology, aims to detect abnormal conditions both in process and hardware. Its mission is detecting pending problems in industrial environment before it developed into severe problems.

mindset detector dashboard

Algorithms

  • Seasonal statistics limits for each measurement

  • Anomalous leaps that don’t violate regulation but hint for pending problems

  • Violating pre-defined engineering rules

  • Recognizing long-term patterns and trends

  • Detection of low-quality data (static parameter, too long fixed values, fixed noise)

  • Unordinary behavior of the I/O cards

  • Abnormal event combination

  • Hazard predefined data combinations

  • Unreasonable noise of group of sensors

  • Similarity of measurements signature to known harmful patterns

MicroEDS has been tested with several standard SBCs such as Raspberry Pi and equivalent products.

All algorithms have a self-learning capability and a default setup which shortens the system’s initial installation.

Operational changes (e.g., change of flow or pressure), cause by changes in the control regime of the monitored process is used by the MicroEDS to filter out false alarms.

A patterns recognition mechanism helps the MicroEDS to identify similarity to known bad or good sequences and to notify the hosting system accordingly.

Main features of microEDS

The MicroEDS is built as a stand-alone product based on a royalty-free operating system. It can communicate with the industrial environment using Modbus/TCP protocol. In parallel, it can deliver reports, datasets and alarms to any configured destination and using standard internet-based protocol.

The MicroEDS builds a mathematical model that describes the relationship between inputs and outputs based on both public and proprietary machine learning algorithms. No knowledge of mathematical modeling is required - models are generated automatically.

To avoid false alarms when your system moves from one state to another, Detector monitors operational changes in process variables.

Adjust for model sensitivity or set target value for false positives and false negatives.

Classify events as Hazard, Non-hazard, Maintenance or Instruments Malfunctioning, to improve model performance.

Remote Access

The MicroEDS can be accessed remotely and securely using VNC protoco Unit comes with a cellular modem, a network cable connection and Wi-Fi.

Cloud backup

Decision Makers LTD supplies automatic backup of the MicroEDS model and data. Based on IBM cloud services an automatic backup services and storage of customer’s data are implemented.

Model Caliberation and Tuning

Using the cloud backup (or ad hoc data transfer), Decision Makers supplies a calibration process for the MicroEDS. This calibration procedure aims to guarantee a low rate of false-positive and false-negative alarms.

We would love to discuss any project you have in mind!

Call Now

©2022. Pinnacle Infosys. All Rights Reserved.