Network Computing is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

IT Meets Operations: 6 Requirements For Industrial IoT

  • You may not think your company is involved in the Internet of Things, but traditional businesses have been using digital technology in industrial capacities for years. Energy, healthcare, transportation, and public safety use remote sensing and monitoring in a multitude of ways. Production lines on manufacturer factory floors are dotted with sensors that monitor output, temperature, defect rates, energy consumption, and other data of vital interest to plant managers.

    This operational technology (OT) has historically been fully segregated from the enterprise information technology (IT) environment. That's largely because OT systems were specifically designed to address factory floor requirements, which are substantially different from the knowledge work requirements fulfilled by IT. Production line management also tends to be somewhat self-contained -- as opposed to knowledge work, which inherently requires interaction across organizations.

    Segregation of OT and IT, however, is rapidly becoming a thing of the past. Factory floor telemetry is of growing interest to business leaders who are looking for new ways to increase operating margins and enhance value propositions to demanding customers. Manufacturers also want to take advantage of IP network connectivity to give plant managers remote access to factory floor data.

    As a result, OT environments are being integrated onto the enterprise network.

    This integration raises a variety of issues for IT. In fact, OT-IT integration can touch every area of IT --including network design, data management, DevOps, and cybersecurity.

    To learn more about IoT and its impact on IT, attend the Internet of Things Summit at Interop Las Vegas this spring. Don't miss out! Register now for Interop, May 2-6, and receive $200 off.

    (Image: zhaojiankang and gogo_b/iStockphoto with modification)

  • Network segmentation

    Generally speaking, OT networks are designed as a single, common network. That may work for the industrial protocols traditionally used in supervisory control and data acquisition (SCADA) systems, but not in mainstream IP networking environments -- where the chattiness of monitoring devices can create problematic broadcast storms.

    Network architects therefore typically need to segment existing networks into separate broadcast domains when they introduce mainstream IP networking when they convert factory floor environments to mainstream TCP/IP protocols as part of a broader OT-IT integration project.

    It's also important to be able to flexibly re-segment the network as traffic patterns change. Many manufacturing environments are designed for adaptive re-tooling in response to shifts in demand. This re-tooling impacts the logical grouping of sensors. So subnetwork configurations have to be modified accordingly, despite the fact that no one from the factory floor team is likely to notify IT's network management staff in advance.

  • Ruggedized devices

    The network equipment that IT typically procures is designed for life in a clean, temperature-controlled data center or, at worst, a wiring closet safely nestled in knowledge worker office space.

    Factory environments are far different. Dust and other particulate matter can post a particular threat to the health of local switches. So can intense vibration and extremes of temperature. In some cases, networking equipment can even be exposed to humidity, vapor, or liquids.

    Network managers need to take appropriate precautions to protect equipment from these threats. Safety measures can include:

    • Use of ruggedized devices specifically engineered for higher tolerance of environmental variables
    • Use of specialized enclosures to insulate devices from vibration, humidity, etc.
    • Longer cable runs that allow network devices to be situated further out of harm's way

    Other risk mitigation measures such as use of uninterruptible power supplies and redundant/failover connectivity are also de rigueur in challenging and operationally critical factory floor environments.

  • Securing the factory floor

    Factory floor networks were historically secure simply because they were "air-gapped" from the rest of the world. Once they are connected to the enterprise network -- which, of course, is connected to the broader Internet -- that gap is gone.

    The potential consequences of a security breach are also very different in OT-IT environments. In IT environments, major concerns include the theft of personally identifiable information (PII) and the interruption of knowledge work through denial of service. Factory floor networks don't typically carry PII, but the consequences of a breach can be very severe if a malicious party can take control of an industrial device.

    Also, when it comes to OT-IT integration, enterprise cybersecurity teams usually have to start from scratch -- since, again, manufacturers have depended almost entirely on air-gapping to keep their digital assets safe.

    Cybersecurity teams therefore need to take appropriate measures to fully protect the OT environment from disruption. These measures can include appropriate network access policies, internal firewalling, and endpoint whitelisting. Logs should also be carefully analyzed and reviewed for signs of anomalous activity.

    (Image: Matej Moderc/iStockphoto)

  • From industrial data to actionable insight

    It's always a challenge to transform raw data into actionable business insight. But manufacturers can face particular challenges as they seek to leverage telemetry from the factory floor. These challenges include:

    • Format conversion. Most industrial control systems -- especially those that have been in place for a while -- have been engineered to support special-purpose data protocols such as Modbus, EthernetIP, and PROFINET. These protocols can vary significantly across devices. So format conversion and normalization are required to make this data meaningful as it is aggregated and sent upstream to the analytic environment.
    • Aggregation and analytics. In disciplines such as CRM, concepts like "customer" and "purchase" tend to be fairly consistent and understandable across different business. Manufacturing metrics, however, can be much more idiosyncratic. This can make the heuristics for understanding manufacturing imperatives such as overall equipment effectiveness (OEE) especially tricky to engineer.
    • Reporting and visualization. Manufacturing management theory has become highly disciplined and formalized in the hundred-plus years since the advent of mass production. So manufacturing managers tend to zero in on very specific parameters regarding line production time, output targets, defect rates, production constraints, and the like. The most effective use of factory floor data requires clear visual reporting and dashboards that give decision-makers clear insight into these classic parameters.
  • Managing the digital value-add

    An interesting phenomenon is occurring as physically manufactured goods begin to contain a digital component. Software itself is now becoming a part of the manufacturing process for everything from jet engines to home thermostats.

    This creates new challenges in the manufacturing process, since manufacturers want to load the latest code on goods as they ship and validate the integrity of on-board firmware and software. Sure, software can still be updated once the product is installed in the field. But manufacturers can't depend entirely on customer-initiated updates. And some basic level of digital integrity is necessary to even ensure that a device can connect to the cloud and update its on-board.

    In some organizations, product code is developed, tested, and managed by a group that is very distinct from mainstream enterprise IT. In others, responsibility for product code devops falls within the IT organization.

    Either way, OT-IT integration requires that the installation of product code onto a manufactured product be managed in much the same way as any other production process -- to ensure quality, to track down root causes of defects discovered in the factory or in the field, to improve output efficiency, and to reduce unit costs.

  • Looking forward

    Most manufacturers are still in the early stages of OT-IT integration. In these early stages, most OT-IT integration initiatives focus on optimizing operational efficiencies and ensuring accurate unit-cost calculations.

    But going forward, IT leaders can expect to see several other trends emerge in the OT-IT arena. These trends include:

    3D printing: This technology, which may be more properly term "additive manufacturing," will significantly impact certain types of production facilities by allowing them to produce components locally on demand, rather than depending on inventory. Conversely, some facilities will do less production because they will simply transmit specs rather than ship manufactured goods.

    Supply chain automation: Over time, supply chain partners are likely to start sharing key data points in order to balance the need to keep production facilities adequately stocked with the need to avoid production bottlenecks. Digital signaling driven by factory-floor data could help automate this kind of inter-company "kanban" discipline.

    Value-based tooling: As production facilities themselves become more adaptable, manufacturers will become better able to make production decisions based on factors besides pure operational efficiency in its original sense. For example, active orders could be mined to determine whether those from an especially high-value customer could be given priority. Rush orders could also be more accurately priced based on known re-tooling costs.

    Of course, every manufacturer's approach to OT-IT integration will be different. A company that produces a wide range of goods for markets with extreme price sensitivity won't have the same objectives as a company that produces premium products bundled with premium services. But every manufacturer will have to up its digital game to compete in today relentlessly evolving global marketplace.

    (Image: chaofann/iStockphoto)