Home > Edge > Manufacturing Edge > Guides > Dell Validated Design for Manufacturing Edge - Design Guide with Cognex > Cross-ISV use cases
The Dell Validated Design for Manufacturing Edge provides innovative integrations that bring machine vision analytics recommendations by Cognex to XMPro for real-time and historical analysis. This process generates recommendations for continuous improvement and process optimization.
A machine vision-based digital twin offering leverages image and associated metadata that is generated from Cognex Vision Pro and Deep Learning applications. Such images and metadata are streamed into the Dell Streaming Data Platform (SDP) for persistence. Ingested data is also automatically indexed for efficient search on demand. Pravega search resident on SDP allows REST API-based searches for customized queries based on specific criteria. An XMPro REST agent, running within a Data Stream, queries SDP for machine vision data and presents the results on the user-defined XMPro Application Dashboard and provides inputs to the user-defined XMPro Recommendations. The XMPro dashboard can also display the image and associated metadata on the dashboard for effective anomaly detection and for verifying the recommendations. By leveraging real-time and historical data-based digital twins, customers can improve manufacturing operations, use actionable insights to make changes to the process, and realize the benefits.
The following are the considerations for system design to ensure effective data ingest and retrieval between Cognex and the XMPro application using SDP as the persistence store.
The ability to ingest and contextualize both machine vision and telemetry data in a digital twin system presents interesting possibilities. Rather than the two data categories residing in siloed systems, value can now be derived from examining the two data types together within XMPro, to deliver previously unavailable recommendations and insights.
As an example, using the case of a Bottling Plant, bottle cap anomalies (e.g. caps missing, cap misalignment) tend to increase as the speed of the bottle conveyor line increases. An XMPro Recommendation rule can be generated to monitor both the rate of cap anomalies (aggregation of machine vision metadata) and the conveyor line speed (telemetry data from deviceWISE). Should both measured values exceed a given threshold, then a Recommendation could be generated to instruct operations to lower the conveyor line speed or indeed execute an automatic action to adjust the speed setpoint.
XMPro interfaces directly with the Dell Streaming Data Platform (SDP) to retrieve images and associated metadata for later display and processing.
The primary XMPro components that are used are the:
Now that the Stream Host is successfully deployed, build the Data Stream required to query the Dell Streaming Data Platform for any new Cognex generated images (and associated metadata) by performing the following steps.
The capability of XMPro to ingest machine vision data from Cognex (using the Dell Streaming Data Platform) enables industry use cases that previously could not be easily realized.
Also, the merging of machine vision data with data from other business systems, within XMPro, enriches the user experience by providing operators with all the data they need, both visual and textual, to manage the plant floor operations from a single application.