Are Automotive Manufacturers Ready for the Internet of Things?
The state of enterprise systems today
With the Internet of Things (IoT) and the ability to manage Big Data, manufacturers can look forward to all sorts of new and improved business processes. Machine data coming from sensors and telemetry will let us know what’s happening in real time so we can make better decisions, faster. We’ll be able to predict future issues and opportunities and use that insight to take the right actions earlier. Manufacturers can be more responsive to customer needs, while also gaining greater profit margins as costs are better managed.
Sounds great – but is this reality yet? On the “crawl, walk, run” spectrum to continuous improvement building on the IoT, most manufacturers today are hardly crawling. Many operate their current “real-time” systems in batch mode with batch planning, batch input of production quantities, batch input of receipts, and so on. When inputs aren’t entered as they occur, the ERP system can’t be relied upon for real-time information for issue analysis or decision making.
Why does batch processing continue?
Many enterprises are stuck in batch mode because their systems struggle with real-time transaction processing. Multihour material requirement planning (MRP) runs are very common. Even regular day-to-day transaction processing can take a long time when product structures are complex. Add to this questionable inventory and bill-of-material accuracy, and it’s no wonder many companies are forced to handle day-to-day manufacturing issues manually. When time comes to make an important decision, the ERP system can’t be trusted for reliable information.
The key question is whether a manufacturer has the ability to operate its enterprise systems in real time. Can receiving, production, and shipping transactions be immediately processed so there is always an accurate perpetual inventory? Is visibility of inbound shipments reliable? Are customer requirements always current? Only by truly operating in real time can the system reveal what is actually happening – and just as important, what isn’t happening when it should be!
What does real time mean?
“Real time” means that the enterprise system is updated with data as things happen. Receipt quantities are visible at the moment items are physically received. Production postings are recorded as product is produced. Shipments are recorded as they leave the dock. Changes in customer requirements are recognized as soon as the customer communicates them. And when possible, plans and schedules are updated as things actually happen that justify or require replanning.
The potential benefits of a truly real-time manufacturing system are enormous. An obvious benefit is being able to access information through reports and displays that represent what is actually happening right now. A not-so-obvious benefit is the ability to monitor for events and conditions that threaten plans, and proactively alert the right people of the possible issue. Think of how much premium freight could be avoided if material shortages were detected a day before the shortage occurred and a materials employee were alerted. The basis of this alert might be a supplier undershipment revealed by automatic shipment notification (ASN) data, or a probable late receipt revealed by a GPS location transmission from the inbound freight carrier.
How many automotive manufacturers are ready to adopt IoT applications so they can reap these benefits? If they’re only “crawling” today, what’s needed for them to walk, then run?
In-memory database makes capitalizing on the IoT practical
Database processing capability is critical. Disk-based databases may be sufficient for collecting enterprise transactions, but they’re many times too slow for real-time transaction processing of complex product structures, the planning or replanning of operations (such as MRP), or the collection, storage, and processing of the huge volumes of sensor data from machines.
It takes an in-memory database to make real-time processing of IoT information possible. Used with data collection technologies such as RFID, barcode scanning, and machine sensors, in-memory processing lets manufacturers operate a real-time system , collecting and analyzing data not only from the plant floor but also from sources outside of the plant – for faster insights, more effective actions, and better day-to-day decisions.