Building a software defined framework for AI automation
While AI is the obvious next step for industrial automation solutions, implementing it is rarely as simple as just dropping an AI model into an existing system and allowing it to run the equipment. Model training and the ability to utilize AI-driven automation means making changes, some minor and some significant, to the way industrial systems are designed and operated.
In a recent podcast, host Spencer Acain is joined by Christopher Schuette, Senior Product Portfolio Manager for Robotics AI at Siemens Digital Industries to explain the move to software defined automation, what that means for industry and AI, as well as what the future holds for AI in factory automation.
Check out the full podcast here or keep reading for some of the highlights of that conversation.
Understanding software defined automation
Traditionally, automation solutions have relied on specialized control equipment and programing languages embedded in systems physically close and directly attached to the hardware they’re operating. While there is nothing wrong with this approach, as automation systems become smarter and employ increasingly complex control algorithms – such as AI – the computational demands make keeping automation on the edge a difficult proposition. At the same time, edge systems have a harder time taking advantage of broad-reaching optimizations and company-wide data strategies.
This is where software defined automation is changing the game. Rather than having dedicated hardware running each instance of a machine on the edge, centralizing and virtualizing that task to a single server allows for far greater compute power to be made available but also for easier management of changes, the ability to both generate and adopt changes based on plant or even company wide data, and the ability to run powerful AI algorithms to control automation systems.
Christopher notes that the move away from traditional PLCs to software defined automation is offering large benefits when it comes to implementing AI enabled robotics, especially when it comes to having access to the hardware and data required to integrate, maintain and run these complex AI models. Beyond these obvious benefits, having a virtual, connected PLC over a traditional hardware solution enables better simulations of proposed robotic systems, ensuring complete validation before implementation.
The future of AI automation
While the shift to software defined automation and AI controlled robotics is already delivering impressive results, that is just the tip of the proverbial iceberg. Christopher explains that, in the near future, VLMs or vision language models, could unlock even greater potential with their ability to directly understand real-world scenes happening in a factory. AI with the ability to directly process images quickly and efficiently, will offer a newer, faster approach to everything from robotics control to defect detection.
Going forward, however, the long-term goal is to achieve a state where machines can be directed by outcome, rather than programmed with individual steps. To reach this state requires the ability for AI to take full control of sensors, actuators and other hardware in a high-frequency scanning, adapting cycle. AI models designed for these scenarios will, hopefully, be able to generalize; building an understanding of the physical world and the processes required to achieve given tasks and adapting in real time based on sensor input.
The ability for automation systems to directly handle tasks from simple instructions rather than detailed programming will radially shift the way automation is handled, especially with the rise of humanoid robotics requiring a more flexible approach to achieve their full value. While achieving this goal is still years away, the groundwork is already being laid and is already showing the promise of more powerful, AI-driven automation in production.
Adopting advanced AI into the factory means a shift in both the way factory automation systems are designed and operated, as well as the way AI models are designed and trained. Together, these changes will alter the landscape of industrial automation from one of fixed robotics performing rigid tasks, to one of flexibility and freedom, with systems capable of being redefined and at will through simple natural language.
Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.
