How it works?
Our software PREKIT enables you to provide advanced, data-driven monitoring & diagnostic tools to your customers.
In 5 Simple Steps
You've already connected your machines and built your data platform? PREKIT runs in multiple modules that seamlessly blend with your existing data architecture. Jump to Algorithm Development.
1
Connect Your Machines
PREKIT Connect
We provide you with a pre-configured LTE Edge Device. Or many.
You mount it in your machine's control cabinet and connect it via ethernet to the machine controller.
Your Data Acquisition Strategy
PREKIT provides you with fine-grained control over...
Data Sharing: Specify which data is logged and whether it is stored and processed on-site or in a central cloud.
Data Model: Enrich data with contextual information by assigning signals to the system elements and process steps they provide information about.
Your Data Management Platform
Based on your configuration, PREKIT generates a data model tailored to your machine.
Providing you with a ready-to-use Data Management Platform.
So you can store and access your data.
Auto-Discovery of Signals
The Edge Device automatically connects with PREKIT and registers the new asset.
PREKIT, via OPC-UA, scans for all sensors & signals available for diagnostics purposes.
2
Manage Your Data
PREKIT implements the DeepFMEA framework. So you can develop, operate and control enhanced monitoring, diagnostics and prognostics algorithms with fine grained control.
Do you want to learn more about how DeepFMEA systematically combines your process know-how with your data to produce more reliable predictions? Read the paper: DeepFMEA Framework
3
Initialize DeepFMEA
Your Use Case
You specify, which critical failure modes you want PREKIT to predict.
And their typical consequences, and how they are managed in the field.
AI-Enhanced Monitoring
PREKIT's probabilistic monitoring algorithm learns how your application and the process it runs behaves, when healthy.
And notifies operators & maintainers, when an anomaly is detected. Together with the information needed to pinpoint its origin faster.
4
Continuously Learning Algorithms
Risk Optimization
PREKIT computes a baseline (economic) risk profile for these failure modes.
Its used to benchmark each detection method based on its impact to reduce your operators' risk.
Enhanced by your Expertise
PREKIT provides you with intuitive visual tools to interpret detected anomalies and assign them to specific failure modes
And compute and analyze meaningful virtual sensor on the data from your entire machine population.
5
Deploy and Operate Your Monitoring & Diagnostics Tool
Tailored Rollout
You maintain fine-grained control over reported metrics & insights for individual target groups (operators, maintainers, ...).
Decide which asset should be monitored.
And choose the reporting channels
(via dashboard, apps, or custom API-integrations).
Fully Automated Operations (MLOps)
PREKIT adapts the diagnostic algorithms to every machine's operating context.
Provides them as a diagnosis tool to all specified customers in the desired way.
Continuously monitors data quality and algorithm performance in operation. And automatically adapts and enhances the algorithms when required.
Want to learn more about ML Ops?
Check Out: MLOps for Industrial Analytics
Provide your customers the insights they need
to proactively maintain and operate your machines.
Leverage proven best practices instead of taking every pitfall.
Build upon mature and robust technology from day one.
Scaling boosts your algorithms' performance instead of melting down your system.
