Matteo Del Balio, Head of Product Marketing at SECO discusses harnessing industrial data for successful network deployments
Digitalisation is transforming manufacturing, with predictive maintenance arguably the most significant application area. In this brave new world, sensor-based IoT systems connect a range of factory floor equipment, collecting real-time data around parameters such as temperature, pressure, vibration, fluid levels, and more. This information is fed into sophisticated data analytics platforms that can identify any statistical anomalies that might identify impending failure. Engineers then make informed decisions based on this actionable insight, promoting the necessary interventions before costly downtime occurs.
Predictive maintenance is just one practical, real-world example of IoT acting as a force for good in industrial environments. And the latest market statistics show that adoption is rising worldwide. Industrial IoT revenues are expected to rise to $238bn in 2024. This upward trajectory will likely continue over the medium term, culminating in a market volume of $454bn by 2029. Subsequently, the number of IoT devices worldwide is forecast to almost double from 15.9 billion in 2023 to more than 32.1 billion by 2030.
The need for effective data fusion
Yet, as the industrial IoT evolves, so the requirement for effective data collection, orchestration and analysis at scale, grows. The ability to merge information from various sensors and systems into meaningful insight for the end user is critical to this requirement. Here, industrial data fusion comes to the fore, representing a significant step forward from traditional data logging to a dynamic, interconnected approach that supports the full potential of IoT.
So, what challenges and opportunities are presented by the need to manage and make sense of data generated by modern industrial systems? And how might the latest solutions, such as advanced Edge computing hardware and comprehensive software suites, help businesses keep pace with the rapid growth of IoT and harness its full transformative potential?
Overcoming data logging challenges
Let’s answer those questions by first looking at data logging challenges today. Industrial manufacturers rely on data logging to monitor and control the parameters required to ensure quality and consistency. These systems combine multicore processors, internal memory, and sensors to log data like temperatures, humidity, pressure, vibrations, voltage, and measurements—key to applications such as predictive maintenance.
Data must then be retrieved and acted on intelligently, requiring orchestration. In this context, data orchestration is an automated process that allows engineers to programmatically create, plan, and monitor data pipelines from industrial edge systems. This automation allows for real-time monitoring of every system and component in an industrial network. It also enables detailed analytics and insights into each machine’s condition – helping avoid downtime.
Phase 1 – Powerful computers at the Edge
However, advancing beyond basic data loggers to more sophisticated AI-ready technologies requires interoperability, integration, and performance to enable a seamless data journey. The first stage in this process occurs at the edge, where next-generation industrial PCs equipped with high-performance processors and the capacity to support GPUs and AI accelerators through modular design reside.
SECO’s Palladio 400 RPL and Palladio 500 RPL are two examples of these types of systems, coming with a powerful 13th Gen Intel Core processor. Designed to operate in rugged environments with temperatures ranging from -40 to 70°C (for the 35W processor SKU), they feature a fanless design, an aluminium alloy chassis, and IP65 ingress protection ensuring resistance to dust, moisture, and vibrations.
Both systems offer versatile modular connectivity, supporting configurations with up to 22 LAN ports (including six 2.5 GbE and eight 1GbE onboard, with additional ports available through add-ons and PCIe). Up to ten of these ports support PoE+ functionality, while Intel vPRO technology enables remote out-of-band system management. For enhanced flexibility, the Palladio 500 RPL features a range of expansion options, including six USB 3.2 Gen 2 ports, two RS-232/422/485 ports, and support for additional interfaces such as PCIe, USB, and SMBus. Plug-in card options include a PCIe 4.0×16 slot, a mPCIe slot, and several M.2 slots.
The systems’ capabilities can be further extended with add-ons that offer high-speed connectivity, enhanced power delivery (60W+), and expanded support for a wide range of cards through M.2 and mPCI3 slots. These add-ons are designed to meet the connectivity needs of Industrial IoT and Industry 4.0 applications.
When data intensive applications require superior AI capabilities, the Palladio 500 RPL excels, especially in configurations featuring Axelera AI’s PCI accelerator card. This AI accelerator integrates the Axelera AI Metis AIPU, which can perform inference operations at up to 856 TOPS and achieve real-time image analysis at an impressive speed of 12,800 FPS – making it ideal for demanding vision applications.
But hardware is only part of the equation. The right software is also needed; this is where solutions such as SECO’s Clea software come in. Clea is a comprehensive platform designed for managing IoT data at scale. Its modular structure allows it to be used either as a complete platform or through individual modules, providing a unified approach to field data handling, device and fleet management, and AI applications.
Phase 2 – Data orchestration in action
With the data collected, the next step in the data journey involves connecting the device to the Clea ecosystem.
Once the connection to Clea is established, data can be orchestrated and transformed into usable information. This requires the use of software that is fully compatible with the data loggers or edge systems. More importantly, it requires a solution that can easily integrate with additional operating systems in the future.
Clea meets this need by acting as the framework for orchestrating and modelling data. Within an Edge device, such as a Palladio 400 RPL or Palladio 500 RPL fanless computer, it automatically categorises, stores, analyses, and optimises data from multiple sensors, seamlessly synchronising across data types and sample rates. It can also auto-generate APIs to enable further integration between systems. Clea also communicates raw or processed data between the Edge device and the Cloud and vice versa.
Moreover, the platform facilitates secure and efficient fleet management for IoT devices, allowing for comprehensive monitoring and control of individual units, device groups, or entire networks. Reliable and ultra-secure, it ensures sensitive data stays protected. It also manages operating system updates and Docker applications, giving more control over the life cycle of IoT devices.
Phase 3 – Supporting end-user outcomes
Finally, in phase 3, the collected data must be turned into usable information that enables/supports end-user outcomes. Clea completes the journey by providing an extensible IoT front-end which allows for full visualisation of devices and data, offering advanced reporting capabilities and fine-grained control over permissions and system access. This interface ensures that the data collected from the entire network is presented in a clear and actionable format, enabling users to make informed decisions.
This ecosystem simplifies sensor fusion, combining data from multiple sensors to identify and isolate failure mechanisms and issues that might otherwise have gone unnoticed.
For instance, combining a high-framerate camera and a vibration sensor might allow engineers to determine that a component on a pump or motor is about to fail. Meanwhile, combining thermal and humidity sensors could identify problems with electrical or steam equipment.
IoT systems that are fit for the future
This three-phase setup provides a framework for collecting, orchestrating, and analysing data at scale. From deploying the hardware needed to collect data, integrating the systems needed to orchestrate and manage it, and implementing the platforms required to use that data to make informed decisions, the latest generation of embedded systems and software suites helps simplify the entire data lifecycle.
Increasingly, the future of industrial automation will require the adoption of such infrastructure to provide engineers with actionable insight. From that comes more efficient, sustainable, and effective industrial operations that are genuinely fit for the future.
Author: Matteo Del Balio, Head of Product Marketing from SECO
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
