Why this article is important:
- The article emphasizes that wireless, edge, mobility and cloud technologies must be adopted as a unified ecosystem, rather than standalone upgrades.
- It reframes traditional ROI discussions around wireless (e.g., cost savings from reduced cabling) to include benefits like improved asset management and reduced downtime, which are central to process efficiency and reliability.
- The coverage of AI facilitating edge-to-cloud transitions, and vice versa, shows process engineers how they can harness real-time data for smarter decision-making and predictive maintenance.
Beyond the physical boundaries of edge, wireless and mobility technologies, the economic drivers behind adoption, funding and deployment must also be considered jointly and strategically.
“Wireless, mobility, edge computing and the cloud aren’t separate entities; they’re interconnected dimensions of a shared economic mindset, converging to tackle larger challenges and drive value creation,” says Naveen Kashyap, chief platform officer of Yokogawa’s cloud promotion business center, cloud hub and innovation activities. “In previous decades, users may have modernized facilities and adopted new technologies simply for their novelty—but that approach is no longer viable.”
For example, the case for wireless from ISA 100 to 5G used to be just seeking a quick return on investment (ROI) due to less cabling, but Kashyap argues it can create value beyond that by also reducing downtime and improving asset management. Since edge and cloud computing are interconnected, Yokogawa reports that artificial intelligence (AI) can facilitate the transition of more edge functions to the cloud when feasible, allowing it to act as a hub for AI execution. Conversely, software models can reach back down into edge systems, serve as points of IT and OT convergence, and execute AI-enabled data gathering and processing at the edge.
“We’re seeing a big onset of hardware at the edge to gather more data for high-level uses,” adds Kashyap. “These technologies are also coming together to improve asset monitoring, energy, optimization, sustainability, supply chains and process dashboards, while providing better user experiences through immersive-reality tools.”
Reinforcement learning assists valves polymer production
Kashyap reports these related technologies and capabilities are also driving into the larger world of industrial autonomy, such as Yokogawa’s vision for its Industrial Automation to Industrial Autonomy (IA2IA) program that’s powered by wireless, edge, mobility, cloud and other technologies, while striving for improved productivity, efficiency and safety. For example, IA2IA acquires data from Yokogawa’s Sushi wireless sensors, robots, control systems, ERP and other sources. It relays this data to a cloud-computing service, which performs analytics, modelling and optimization. IA2IA and the cloud also creating reports and dashboards for use cases, such as digital twins, energy and production reconciliation, and supply chain planning and optimization based on the joint expertise shared by Yokogawa and its KBC subsidiary.
Similarly, Yokogawa recently used data from equally disparate edge and wireless sources to help ENEOS Materials Corp. achieve autonomous control aided by AI on a polymer production column at its plant in Yokkaichi, Japan. The AI it implemented was factorial kernel dynamic policy programming (FKDDP), which is a reinforcement learning-based AI algorithm that was jointly developed by Yokogawa and the Nara Institute of Science and Technology. FKDDP helped control valves and liquid levels, and responded to ambient temperature changes, which had been difficult for the plant’s operators, DCS and advanced process control (APC) system. Over a year, FKDDP and its self-learning AI model we’re able to control the valves efficiently, making ENEOS the first company to achieve long-term, AI-based autonomous control, according to Yokogawa.
“Process control with existing cloud and edge devices can be quite challenging," explains Kashyap. "In this case, FKDDP seamlessly integrates AI with the automation stack to enable autonomous control.
