Custom Meta Tags

Hero Banner

Your Perfect Partners for IIOT

Main Title

The Design Engineer's Guide to the Industrial Internet of Things

body1-intro

In this guide, we take a look at how the IIoT is impacting industrial automation and manufacturing, maintenance and analytics, quality management, and logistics and supply chain. The areas of sensing and connectivity, specifically wireless and wired, will also be reviewed in this context.

The IIoT continues to focus on communication and connectivity. Of course, some additional technologies, such as 3D printing and other additive manufacturing systems, along with artificial intelligence (AI), will contribute to the spread of industrial IoT. However, these will depend on data-sharing in order to maximize their impact.

A key focus here is how existing industrial process can be made more intelligent and Internet-connected. This begins at the sensor, moves up into the cloud, and then back to the actuator (Figure 1, below). A few key terms to note:

Sensors – these measure physical properties, such as temperature and pressure, and are often capable of measuring several physical properties at once. They typically pass their data in analogue or, increasingly, in digital format to either gateways or edge computers.
Actuators – these turn control information from gateways or edge computers into various types of motion (e.g. linear, rotational). They provide motion for robotic arms, conveyor belts, and dosing machines, to name a few.
Gateways – an aggregator/distributor of data, typically linking one networking technology (e.g. HART or Bluetooth) with another (e.g. Industrial Ethernet).
Edge computing – a local computing resource whose close proximity to both sensors and actuators enables the lowest latency, and real-time control of equipment based upon incoming sensor data.
Cloud computing – an off-site computing resource, typically hosted by a third-party provider. The computing power and storage can be expanded or reduced “elastically” as required. Usually, it performs analysis on many datasets, possibly using AI, and provides methods for real-time monitoring by operators of industrial processes.

Fog computing – provides an on-site computing and storage resource to balance between the limitations in performance and storage capacity of edge computing and the relatively long latencies of cloud computing resource.

Figure 1: The IIoT will link together a wide range of disparate sensor and actuator technologies with edge, fog and cloud computing resource

body2-content

Sensing technology that meets the needs of harsh industrial environments will be essential in ensuring that engineers can realize their Industry 4.0 ideas. In a recent survey by TE Connectivity, 29% of engineers, including those with an interest in IIoT, indicated that the ability to capture different kinds of data would be critical to their application. Furthermore, 26% felt gathering more data faster would be a significant consideration.

This will also require advances in industrial networking technology to enable more compact, intelligent sensors to monitor processes and machines. Fifth-generation cellular mobile networks, promising low-latency wireless networking, are considered to be a key element here. The wide range of services and software, from cloud to edge, that provide the glue to link these interconnected sensors, machines and computers together, will also be a major contributor.

body3-expandable-content

How will the IIoT impact industrial automation?

The depth of insight provided by the Industrial IoT will help improve manufacturing across the board, from reducing labor costs to lowering design engineering spending. Even in countries where human labor costs are perceived to be low, such as Asia, labor costs are actually rising. By networking machines, their sensors, and production lines, and enabling powerful cloud-based computing to participate in some of the decision making, industrial automation can be advanced further by the Internet of Things. Over time, human resources can move from low value tasks to more cognitively interesting activities, such as managing and maintaining manufacturing plants and the implementation of such systems.

Industry 4.0 trends and the future of manufacturing

With more and more information about manufacturing processes being stored in the cloud, thanks to networked intelligent sensors, designers will be able to factor known manufacturing challenges into their designs. Sharing of such data will contribute to best practice techniques and approaches, in turn leading to efficiency in material usage and improvements in product quality.

Figure 2: AGVs are already an integral part of some manufacturing plants, with AGVs maneuvering work pieces between production islands (Source: Siemens Ingenuity for Life community thread

Further applications of the Industrial IoT in manufacturing

What other use cases does the Industrial IoT open up?

Ultimately, we could get to the point where customers can order products in varying batch sizes and meet all required customizations via the web. These will be automatically fed into the factory and the desired product delivered with almost no human interaction. This involves moving away from linear conveyor-belt manufacturing lines to islands where production steps are realized (Figure 2, right). Automated guided vehicles (AGV) are used to move the work piece from island to island, with the number of visits to each island and the order of visits dependent upon the precise customer order. This approach is already being utilized in some semiconductor manufacturing facilities¹.

What will the IoT mean for predictive maintenance and analytics?

Perhaps the most significant potential change presented by Industry 4.0 is the move from preventative to predictive maintenance and analytics. Rather than calculating the moment at which equipment is likely to fail and then shutting it down for maintenance prior to this date, manufacturing equipment would be able to register a pending maintenance need itself.

Sensors will play a crucial role in this new world of predictive maintenance. Sensors are now available that can determine the state of motor bearings using fast-Fourier transforms (FFT) to detect unusual vibration frequencies² in manufacturing equipment (Figure 3). Edge-based artificial intelligence analytics, coupled with sensors that consider vibration together with other factors (such as temperature, load, current draw and the current activity), could completely change how maintenance is approached in manufacturing environments.

If such insights are shared with the suppliers of industrial machines, such as manufacturers of motors, ovens and conveyors, they could result in improvements that benefit all users.

Figure 3: Performing an FFT on vibration sensor data is one approach to detecting wear in motor bearings for predictive maintenance

What quality management benefits can the IIoT deliver?

Often quality issues are not detected until a batch of completed parts are being assembled. By this time, valuable material and production time has been wasted. Under Industry 4.0 an increase in the use of sensors and the utilisation of cloud-based AI to evaluate disparate data sets will help to detect anomalies that, today, are determined post-event. High integration of sensors, such as combining temperature, humidity and vibration into a single device, will also broaden the base of quality management data available for analysis.

Placing AI at the edge, such as in cameras, will provide intelligent sensing capability for a broader range of applications. In combination with the island manufacturing approach described earlier, such cameras could handle the visual inspection demands of a wide variety of customisations for a standard work piece.

Which aspects of the Industrial IoT will improve logistics and the supply chain?

Many businesses today still make use of best-guess demand to develop their forecasts. With a fully cloud-connected manufacturing plant, it is theoretically possible that incoming orders could be planned-to-order. Just-in-time material delivery would be managed based upon orders for the day, with goods outwards optimized in the same manner. These forms of supply chain optimisation could reduce the amount of raw material held in goods-inward, as well as finished product in the warehouse.

Companies such as Ocado are already demonstrating highly automated warehouses for retail products, but the same principles apply to many industrial manufacturing facilities³,⁴. Here, 4G cellular networks are used to control autonomous robotic pickers from a control centre akin to air traffic control (Figure 4, right). AGVs can also play a role in the first and last-mile delivery of goods and parts. Companies such as Starship Technologies have been trialling robotic delivery services that could pave their way for integration into the wider supply chain⁵.

Figure 4: AGVs working autonomously to prepare delivery of items for shipment (Source: https://risnews.com/krogers-weapons-grocery-home-delivery-war)

Feeding the IIoT with sensing information

While early industrial sensors relied upon discrete components to implement their measurement function, many today can rely upon highly-integrated, low-cost semiconductor technology. Devices, a few millimeters on each side, are capable of integrating pressure, vibration, temperature sensing and more into a single piece of silicon. Thanks to microelectromechanical system (MEMS) techniques, sensors can be made up of vibrating elements and cantilevers that have features measuring 1µm to 100µm in size.

Alongside these minute sensors, engineers integrate all of the front-end filtering and analogue-to-digital conversion required to deliver precise and calibrated output data. Such compact, precise solutions can enable highly accurate positioning of work pieces by robots beyond what is implemented today.

Industrial sensor manufacturers are also provided with more freedom, enabling them to pack more into a standard housing than ever before. This could include wireless technology, such as Bluetooth, allowing data to be shared with a maintenance engineer’s smartphone, or even a display to ease process analysis when at the equipment (Figure 5, left).

Protection of remote equipment can also be a security challenge, especially in the age of networked-everything. Tamper detection sensors (TDS) can be utilized to not only detect undesired access to remote systems and hardware, but also to ensure that any useful information, such as encryption keys or access information, is automatically erased, rendering the hardware useless.

A lot of legacy equipment in the field today is built upon fieldbus technology that fulfilled the needs and approaches of the day. Today’s industrial sensors continue to support analogue output standards such as 0-5V, 0-10V and 4-20mA, requiring a cable per sensor to connect it back to a PLC for processing. Digital networking technology, both wired and wireless, provides methods that enable the connection of several sensors to a single PLC. This, in turn, provides more freedom and less cost when increasing the quantity of sensing used. It can also be easily shared with other systems and the cloud.

Figure 5: The M5800 pressure sensor from TE Connectivity includes its own display

Challenges of wired connectivity for IIoT

Sensors are often connected over extremely longs runs of cable and may be part of a safety critical system. Industrial settings are also harsh, both with respect to environmental factors, such as temperature, moisture and humidity, as well as electrical interference, generated by switching circuits, welding tools, and motors. Therefore, both connectors and cables need to be not only electrically sound but also robust to moisture, chemicals, and potentially high-pressure liquid cleaning processes.

Connectors must also be protected against unwanted disconnection, either as a result of vibration or accidental force applied to the attached cable. Many established connection technologies, such as Ethernet and USB, were designed for home or office environments, where frustration is the worst outcome of an unexpected disconnection. Industrialized connectors prevent such outcomes by implementing locking mechanisms that increase the retention force of the industry standard connector, or by implementing ingress protection, known as IP rating, which stops dust and liquid (Figure 6, right).

In addition to ruggedized USB or RJ45 Industrial Ethernet connectors, there are a range of standardized industrial multi-pin connectors. These are often designed to support the transfer of both digital and analogue signals along with power to the attached sensor. It is critical to understand where such M8, M12, Mini I/O and D-Sub connectors are used to ensure hardware is compatible with existing systems.

Figure 6: Industrial applications typically demand a more robust connector that resists dust and moisture ingress

Approaches to wireless IIoT applications

With AGVs already in production environments, wireless is increasingly becoming a topic that industrial engineers need to address. Like wired connectivity, many technologies, such as 4G cellular networks and IEEE 802.11 WLAN, have been driven by consumer and office needs. Additionally, the bandwidth for the upcoming 5G standard is currently being allocated and is seen as a key IIoT enabler. Wireless, low-power asset tags, which can track items from delivery, through the factory, until they leave again, or even provide asset management for manufacturing equipment, are also in consideration.

Here, innovative antenna solutions will be required that are both optimal in their radio functionality and can withstand harsh environmental conditions. At their simplest, a single element antenna, possibly as a chip-antenna or designed into a flexible printed circuit (FPC), is one approach (Figure 7). However, radio solutions that use multiple-input, multiple-output (MiMo) antennae are becoming an essential element of wireless connectivity.

Regardless of the approach taken, many engineers need support and guidance when tackling the RF side of their application. Custom antennae can often be the best approach, allowing the designer to take into consideration the materials used, the environment, and the object upon which the antenna is mounted.

Figure 7: Custom antenna solutions can be integrated into FPC or applied to 3D mouldings to meet the needs of the application

Sensitive electronics also need protection from electromagnetic interference (EMI). Board level shielding with a stamped one- or two-piece construction method, may also be required, both to restrict the impact of interference from neighbouring equipment, as well as to reduce interference generated by the electronic system itself.

body4-summary-and-references

Summary

Like the industrial revolutions before it, Industry 4.0 is not a step change. Rather, it is a progressive development toward new, more efficient approaches to industrial manufacturing. IIoT can be challenging to nail down at times, as many ideas remain conceptual in nature. However, this is changing as industry realizes that existing continuous improvement processes are at the point of diminishing returns. Fully automated warehouses already operate on a daily basis providing order fulfillment, while AGVs roam the streets of selected cities, carefully gauging the unforeseen challenges of introducing such technology to the public.

The optimal approach to IIoT is to openly discuss potential ideas, even when they are not fully formed or when they contain challenges that are currently unsolved. Suppliers and partners can often draw on experiences and technologies to support their customers in such cases. This can be as simple as offering a suitable connector, or as involved as in-depth consultancy on antenna selection and design, or even optimal approaches to EMI challenges.