MaDaM, the acronym of Measurement Data Management, is the organizational tool of AMS and the first step of its big-test-data toolchain – a complete cluster/data-mining solution. In Version 2, MaDaM comes with Elasticsearch indexing technology and a modern web interface. The complete application offers horizontal scalability without limits. Fast page rendering as well as quick response times can be achieved, because it is built on the modern JavaScript-based framework, ReactJS.
For the update 2.2, AMS developed new features, such as Haystack for visual searching by clicking on the relevant bars in the graph, a leading wildcard search for encapsulated contents as well as the control of analyzing multiple tests. This analysis is done via jBEAMTM, AMS’s data analysis and visualization tool. “It is our main goal to provide optimal conditions for engineers to analyze their big data sets dynamically, depending on the daily changing issues. New powerful, engineering related, software is required to solve this challenge,” says Dr Bernhard Sünder, managing director of AMS, referring to the increasing amount of measured data these days.
“More than 30 MaDaM systems are in use globally at test facilities, including Daimler and Patac,” continues Sünder. “The combination of MaDaM’s effective data management together with the analysis and visualization tool, jBEAM, provides a complete solution for local and distributed test data management and reporting in all fields of R&D: crash tests, test drives, NVH, and so on.”
Collecting test data over several years requires the evaluation of probably more than 100,000 tests. It is therefore necessary to run parallel analyses with jBEAM-Cluster to detect certain events and related parameters within this bulk of data in a short time. This aggregated list can be analyzed with Data Mining algorithms in a second step to find reasons of malfunctions or specific behaviors.
With MaDaM’s platform independence, engineers around the world are capable of managing colossal amount of information, whether they are using Microsoft or Unix systems. The new features of MaDaM 2.2 also contribute to the convenience of the user-friendly interface and workflow, so that the engineer will be able to perform Data Mining processes without the need of a data scientist let alone a software developer.