ABI Research, a market-foresight advisory firm providing strategic guidance on the most compelling transformative technologies, has identified several trends and use cases whereby edge intelligence augments cloud platforms in the manufacturing sector. Some of the key findings include:
- Several vendors, such as AWS, Azure, FogHorn Systems, Software AG, SWIM.AI, and Telit, have already productized edge to cloud closed-loop machine learning.
- Enterprises in automotive, electronics, oil & gas, and steel manufacturing, among several other industries, have already implemented and seen ROI from edge solutions.
- As the amount of custom code required to deploy new solutions on factory floors drops, data and analytic service revenue growth in smart manufacturing will accelerate to reach a global total of US$25.6 billion in 2026, led by the USA with US$5.3 billion in revenues.
“If smart manufacturing vendors hope to fulfill the potential of their solutions and platforms for digital factories, they must build environments where apps can deliver immediate results at the edge with stream processing and integrate with the cloud,” says Pierce Owen, Principal Analyst at ABI Research. “Cloud platforms can integrate factory data with the supply chain and the rest of the enterprise, train machine learning models, and help scale other transformative technologies, but they need edge intelligence to truly integrate with OT, lower Total Cost of Ownership (TCO), and stay financially viable for customers with limited IT resources and increasing amounts of data.”
Use cases that leverage edge solutions include networking machines from multiple OEMs with proprietary protocols, quality control for automotive paint shops and windshield glazing, executing machine learning models on-premise, predictive maintenance for almost any manufacturing equipment and automatically orchestrating setup of production lines for electronics contract manufacturers. Telit deviceWISE provides protocol translation, edge intelligence, and code-free app development and integration. FogHorn Systems and SWIM.AI also offer edge intelligence. PTC’s Kepware connects and integrates equipment with protocol translation while its ThingWorx platform can run in its entirety on HPE Edgeline systems. Software AG provides Cumulocity IoT, which covers all its device connectivity, data acquisition and management software, including CEP in its Apama product, closed-loop predictive analytics in its Zementis product, code-free ML analytics and alerts in its TrendMiner product and integration through its webMethods product.
In the past, networking OT equipment from different OEMs often meant continuously working with custom code, which resulted in immense costs and time demands on the IT departments. Now, many edge solution vendors have started to offer code-free logic configuration and app development, significantly speeding up deployment and scalability of custom apps on the factory floor. As more vendors offer and deploy code-free app development and logic configuration at the edge, smart manufacturing solutions will grow faster, requiring fewer professional services.
“As in many other verticals, the next step for smart manufacturing will demand faster deployment of AI and machine learning. More low-code or code-free logic configuration and app development, plus edge to cloud closed-loop machine learning whereby models train in the cloud, execute at the edge and collect more data for continuous improvement, will make this possible. Fortunately, several vendors have already proven that they can provide both,” concludes Owen.
These findings are from ABI Research’s Balancing Edge and Cloud in the Digital Factory report. This report is part of the company’s service which includes research, data, and Executive Foresights.