Automotive manufacturing environments generate data from the assembly line all the way through to quality control, inventory, equipment, and materials. Connecting this information supports manufacturing operations' personnel to understand what's going on in their environment, but how can manufacturing teams effectively leverage connected environments and operations?
Artificial Intelligence of Things (AIoT) combines connected things along with data intelligence to improve operations decision-making
This article focuses on several technical areas related to connected automotive manufacturing.
- Production visibility and work-in-progress
Vehicles and their parts move through many steps of production before hitting the road. Understanding where production activities are at any given time can be a challenge without the ability to see what's really going on. A connected operations application can ingest and associate relevant data from manufacturing operations management systems, tracking systems, and factory applications to create a better view of what's going on with production activities.
Implementation considerations
How to structure and identify vehicles, parts, and production orders
What production events to capture
How to associate WIP tracking information with production applications and data
What production visibility information to expose to production personnel
This is about associating the right information with production activities, rather than some generic WIP dashboard
- VIN Genealogy/Manufacturing traceability
Understanding what happened to, or with, a particular vehicle or serial number throughout its production life can have value. This might include relevant manufacturing activities for that asset, depending on how that system is architected.
Traceability System Considerations
Identity (associations): How genealogy information for particular vehicles or numbers is associated and identified
Events: What manufacturing and quality-related events for vehicles or serial numbers are relevant
Relational Integrity: How component and quality events or data are associated with the vehicle or serial number they apply to
Access/Usage: Making sure this manufacturing intelligence is accessible throughout the relevant production applications
Genealogy information is only going to be as good as the integrity of your source data and event definitions. Simply capturing more events is not necessarily going to enable traceability.
- Automotive Inventory visibility and production coordination
Production operations rely on materials being available. Inventory systems and production systems might have relevant information or a different perspective on that availability.
By connecting and associating these two sets of information, the manufacturing operations team can better understand materials availability, consumption, WIP, and so forth in a connected view that adds value to their operations. Particular considerations here might include:
How often inventory updates occur and how it synchronizes with production events
Consistency of identifiers across systems
Events captured by each system
How exceptions are surfaced to the appropriate teams or manufacturing applications
RTLS for awareness
Depending on how it's deployed, a Real Time Locating System can provide position information for a variety of factory objects - vehicles or vehicle parts, people, materials, or other WIP. RTLS can leverage different access technologies depending on environment and needs. This RTLS data can become associated with production order information or accessed by factory applications seeking awareness of particular events.
Before jumping into an RTLS deployment, manufacturers might consider what objects to track, required accuracy, environment considerations, and how to make that data available to relevant factory applications.
These considerations are about turning raw position data into actionable manufacturing intelligence for production teams.
- Bringing it all together: data contract and system interfacing
Inventory data, production telemetry, quality events, and location data from RTLS are all coming from different sources. Connecting them requires more than just a pretty UI; it also requires a common contract for exchange of information.
Some important considerations here include
The data contract - field names, meanings, value domains, and other metadata needed for applications to consume this data
Temporal aspects of event processing, such as handling out-of-order and duplicate events
Interface design/data access requirements needed for each application to allow them to consume this information
Data quality considerations (gaps, NULLs, out-of-sync or incorrect identifiers)
Permissions/access control: policy and data governance considerations for who can do what with the data
Operations and maintenance: availability, freshness, and failure-management considerations
All of these are important pieces for helping turn your connected manufacturing data into actionable manufacturing intelligence within your AIoT application.
- Starting with a use case
This is about starting small - identifying a particular application need that could benefit from an AIoT or connected application. This might include production visibility for work in progress, genealogy with VIN numbers, or inventory visibility for materials. The idea is to identify one thing that you think would help and try it out. If you make it this far, you can always build other use cases on top of it.
Now What?
AIoT can help you turn disparate sources of manufacturing data and operations information into something meaningful for your team. But how much does it really help? It really depends on the quality of your data, system interfaces, and whether you know what you're trying to accomplish with the information. If you're looking to connect your factory applications, leverage this information more effectively, and improve operations, consider exploring AIoT opportunities for connected automotive factories.
For additional information about AIoT opportunities for intelligent connected automotive factories.
Visit : "OEM Nex AI"
The essential engineering question addressed by this article is, "How can manufacturing systems combine isolated operational events?"
