ARC Advisory Group attended the IoT Solutions World Conference (IoTSWC) held in Barcelona, Spain last month. Our general comments from last year still hold true. The conference includes a fair with over 18,000 visitors and a congress with over 200 presentations and 5,000 participants. Furthermore, the event presents an area for IIC testbed demonstrations, and an area for promising startups, another growing event. IoTSWC is held in parallel with the Blockchain Solutions World and the AI & Cognitive Systems Forum; together they form Barcelona Industry Week. The congress had nine thematic tracks this year, including IoT enabling technologies, blockchain and artificial intelligence, connected transport, manufacturing, energy, and utilities.
IoT Use Cases
Modernize OT to Leverage AI
We attended a lecture by Nutanix and the Hardis Group about modernizing Operational Technology (OT) to leverage Artificial Intelligence (AI). The Hardis Group provides supply chain and warehouse solutions using AI, using camera and sensor information to analyze flows and provide dashboards to operators providing statuses and highlighting issues to resolve. Nutanix explained that virtualization of the connectivity to OT devices, for example Kepware, Wonderware or Litmus Automation, helps to stream data to a single bus and overcomes the silos of disconnected datalakes created by separate solutions. Nutanix suggests using simple Python code to cleanse the data. On that base, AI applications can be developed in the cloud and deployed in the cloud or at the Edge. Nutanix’ Edge platform enables deploying models in containers, thereby creating scalability through hardware independence. Models can also be updated during industrial production at the Edge.
Normalized Data to Access Legacy Information
We also attended a presentation by Danfoss and Omnio. Danfoss produces drives and frequency converters for electric motors and have an installed base of 20-25 Million drives produced in the last fifty years, most of which are not native IoT devices. Nevertheless, Danfoss is interested in providing analysis of their drives’ behavior and performance to create opportunities for optimization, both for their clients’, as for their own operations. As a test case for the new approach, Danfoss took HVAC systems in their manufacturing operations where different types of drives are used, combined with sensor information. With a variety of protocols, parameter naming conventions, units of measure and scales, and time synchronization issues, it turned out to be a lot of work to get access to the data.
Omnio, a provider of device onboarding solutions for industrial IoT, built an “OT middleware” that provides connectivity to fieldbus and other protocols, as well as the normalization of the data model, which was designed with the client. In this case, over thirty parameters were considered significant to understand the operation. Once the data were in a normalized format, Danfoss worked with Accenture to develop multivariate analyses and get an understanding of process behavior. What is very useful in this application, is the realization that a drive, with its built-in sensors and computation, can be used as a sensor, an edge computer, or a controller. This way it can provide a wealth of information, accessible through the traditional industrial control system via the fieldbus. No additional sensors and wiring are required to provide hundreds of data points. Moreover, if an additional sensor, for example to capture vibration of the motor, would be needed, it can be powered by the drive, and the measurements can be provided via the analog input of the drive through the fieldbus. This approach would be very attractive for many sectors in the process industries, where most drives are connected to control systems via fieldbus.
It will address test beds, an in-depth meeting with FIWARE on what the context broker really is and a brief introduction to GAIA-X thé EUROPEAN Cloud initiative