Skip to main contentIBM Garage Event-Driven Reference Architecture - Reefer Container Shipment reference implementation

Anomalous Container Scoring Microservice

The Anomalous Container Scoring microservice consumes telemetry events for reefer containers and applies a predictive scoring model to determine whether or not the specific shipping container is in need of maintenance or not. The scoring service uses an analytics scoring model build using machine learning techniques, and depending upon deployment options, can be serialized and loaded into memory.

Overview

Description: This microservice is responsible for listening to the Reefer Telemetry Topic topic where the IOT sensor devices of the Reefer Containers will be sending their telemetry data to. These telemetry events will be read and used to contact a container anomaly prediction service based on Watson Machine Learning, hosted on the IBM Cloud. Depending on whether the prediction service predicts a container anomaly, this Telemetry microservice will send a Container Anomaly Event to the Containers Topic for the Containers microservice to handle the shipping goods spoilage.

This microservice has been implemented using the latest Reactive Messaging feature of MicroProfile 3.0. running on the OpenLiberty server.

Telemetry diagram

Github repository: refarch-reefer-ml

Folder: scoring-mp

Kafka topics consumed from:

Kafka topics produced to:

Events reacted to:

Events produced:

EDA Patterns implemented:

Build

Run

Usage Details

REST APIs