The global site of the UK's leading magazine for automation, motion engineering and power transmission
18 June, 2024

Twitter link

AI module sits in a controller, looking for application problems

16 November, 2017

Rockwell Automation has developed an artificial intelligence (AI) module that fits inside an industrial controller, “learns” the application that the controller is managing, and builds a model of it. It then monitors the operation continuously, looking for anomalies. If it spots a problem, it can trigger an alarm on an HMI screen or dashboard. Future versions will go beyond diagnostics to guide users on how to remedy the problem, or adjust system parameters automatically to fix the problem without human intervention.

The development – known as Project Sherlock and unveiled at Rockwell’s annual Automation Fair, taking place in Houston – has been running in pilot installations for the past 18 months, and is due to go on sale in mid-2018.

Once plugged into the backplane of a Logix controller, the module uses novel physics-based modelling to “learn” the application that the controller is managing. It scours controller tags to identify the application, or allows users to choose what they would like to be modelled by choosing inputs and outputs via an add-on instruction. The AI algorithm then learns from the stream of data passing through the controller to build a model. It can do this in a matter of minutes. Vast quantities of historical data are not needed, nor must the data ever leave the automation layer.

“Project Sherlock brings industrial producers amazingly smart analytics in a package that is easy to implement,” explains Jonathan Wise, platform leader for Rockwell’s Control and Visualisation business. “As our customers undergo digital transformation – using production data to help improve business outcomes – they can’t wait for expert-driven analytics.

“Even if there were enough industrial data scientists out there,” he continues, “not every company has the time or funds to employ them. This machine-learning tool creates powerful analytics from your automation infrastructure, painlessly – delivering value moments after it’s dropped in the Logix backplane.”

Rockwell claims that Project Sherlock’s diagnostics will cut false-positive alarms “drastically” compared to other AI systems due to the physics-based modelling and its foundation in industrial applications. For example, the AI can tell if a shift in boiler temperature is related to a benign change in upstream operations or to an abnormality that requires correction.

The initial version of Project Sherlock AI will include ready-to-use templates for boiler, pump and chiller operations. Users can model additional applications with guided configuration.

The Project Sherlock module places artificial intelligence inside industrial controllers

Communications with the module are prioritised by the controller, so users can choose how much data is sent and the communication intervals. The module does not add to the host controller’s CPU load or to network traffic.

The AI engine is part of a larger, expanding ecosystem of analytics offerings from Rockwell Automation with applications ranging from devices, machines and systems on the plant floor, to the complete enterprise. The company’s developers are building connections so that users of FactoryTalk Analytics for Devices tools will be able to interface with Project Sherlock AI via its Shelby chatbot and action cards. Analytics from Project Sherlock AI will be integrated easily into the FactoryTalk Analytics platform to integrate plant-floor data into business intelligence strategies.

•  Rockwell Automation has recently invested in The Hive, a Silicon Valley innovation fund and co-creation studio which identifies AI-based opportunities with growth potential, and co-creates, funds, and launches startups with operational and technology support. Rockwell says the aim of the investment is to gain access to innovators and technology start-ups with a focus on applications of AI to industrial automation. Its goals include co-creating to solve customer problems, accelerating innovation, and identifying new emerging technologies that can help its manufacturing customers to improve business performance by bridging the gaps between plant-floor and higher-level information systems.

“Smart manufacturing requires the use of new and disruptive technologies such as AI to create the future industrial plants and supply networks that are flexible, efficient, responsive, and secure. AI can help manufacturers unlock data, contextualise it, and take action,” says Elik Fooks, senior vice-president for corporate development at Rockwell Automation. “We continue to create partnerships with leading innovators, such as this one with The Hive, to further advance The Connected Enterprise, our vision for realising unprecedented industrial productivity from the integration of plant and enterprise operations.”

  • To view a digital copy of the latest issue of Drives & Controls, click here.

    To visit the digital library of past issues, click here

    To subscribe to the magazine, click here



"Do you think that robots create or destroy jobs?"



Most Read Articles