Capabilities demonstrated will be from the new Predictive Maintenance Toolbox in MATLAB.

Attendees will be able to interact with the presenters and see how they can evaluate different techniques and modeling methods for developing their algorithms. If possible, we will try and bring a small hardware setup with a servo motor that can be placed under different fault conditions. The data from these faulty conditions can then be analyzed in MATLAB.

The demo will begin with the explanation of QSI’s multi-signal modeling concept, the core of QSI’s Model Based Systems Engineering (MBSE) approach towards Field Maintenance and Remote Support. QSI will demonstrate its graphical modeling tool, TEAMS-Designer, and show how a cause-effect model of a system is used to improve overall reliability and generate optimal diagnostic strategies. Various analysis capabilities such as Testability Analysis, FMECA (Failure Mode, Effects and Criticality Analysis), FTA (Fault Tree Analysis), etc. will be shown.

Next, we will be showing how the same TEAMS model of the system can be used to conduct Real-time Health Monitoring and Guided Troubleshooting through the use of QSI’s enterprise-grade software, TEAMS-RDS (Remote Diagnosis Server).

The tech demo will familiarize users with multi-signal modeling fundamentals, and show how they can create system representations from an MBSE perspective. Additionally, by utilizing features such as “Design For Testability (DFT)” and “Test Recommendations”, we will demonstrate how to improve the “testability” related qualities of the model.

The tech demo will also provide background information about maintenance concepts such as Diagnose Before Dispatch (DBD), Condition Based Maintenance (CBM) – and how it differs from Reactive/Unscheduled Maintenance.

Asset Answers helps answer many of these questions. GE Asset Answers aggregates millions of work history and equipment records from the CMMS/EAM (Computerized Maintenance Management Systems/Enterprise Asset Management) from industrial facilities around the world. There are millions of historical maintenance activities recorded on millions of assets across different verticals in the Asset Answers database, which can be used to characterize different maintenance strategies, different commonly occurring failures, and characterize failure patterns and their cost. This data is anonymized and made available to subscribers who can compare themselves against peer data. Reliability engineers can dispense with spending a bulk of their time on data processing and aggregation, and have data at their finger-tips, ready for analysis as soon as they login to Asset Answers. By incorporating data into decision making, AA along with APM, enables an organization to go from reactive (fire-fighting) to proactive (predicting and preventing failures).

The availability of more data is an enabler for the data scientist that allows new approaches when constructing AI based algorithms or, during the training phase, achieve higher accuracy. Challenges arise when such data is found across multiple organizations when access and usage needs to be arranged. The Digital Data Marketplace is an approach where organizations, as members of a consortium, recognize a common benefit to make their data available to enable development and training of algorithms, however access and usage needs to be arranged in a trusted, fair and economic way. In our Tech Demo, we will show how the SAE ITC ExchangeWell consortium initiative can facilitate organizations that like to achieve common benefits no single organization could achieve on its own, whilst addressing member concerns regarding data value, confidentiality, intellectual property, and more in a trusted way. We will demonstrate how a consortium driven marketplace operates, where data owners and algorithm developers register their assets and subsequently negotiate visibility, access and usage of their data assets. Novel data science processing concepts, such as Federated Analytics using a consortium infrastructure helps protect data assets. We will explain how blockchain and smart contracts is a key enabler for the underlying operation and orchestration of data science workflows resulting from supply and demand – members negotiations.