This piece was originally published in the August 2016 issue of electroindustry.
Robin Duke-Woolley, Founder and CEO, Beecham Research
In offering transformative rather than incremental change, the Internet of Things (IoT) can reshape whole industries—not least of which is manufacturing.
According to Accenture, the IoT will add trillions of dollars to the global economy over the next 15 years, and a study by Cisco found that the manufacturing sector will drive 34 percent of total IoT value in the overall economy through 2024. Gains are already being realized: a survey by Tata Consultancy Services indicated that industrial manufacturing companies that invested in IoT technologies saw, on average, a 28.5 percent increase in revenue between 2013 and 2014. The same study revealed that those companies expected a 27.1 percent increase in revenue over the next three years.
Behind the numbers lies the reason for these gains: the IoT enables business transformation through connectivity to internet protocol (IP) networks and resources. It offers a way to change the basis of competition, allowing companies to innovate, realize gains in productivity and efficiency, and grow revenues. In this newest industrial age, customers expect more than a traditional, fixed-end product; they are looking for products that are customizable to their needs and improve over time. The IoT allows companies to offer such products and enables them to offer services that generate new revenue flows and surpass customer expectations. Businesses that ignore the full impact of the IoT, therefore, do so at their peril.
The IoT goes far beyond simply monitoring operations. Rather, it can provide better insights, suggest methods of improvement, and enable greater efficiency by combining sensor inputs with analytics and applications connected to powerful platforms using fast, reliable connectivity methods. Two key components differentiate the IoT from remote monitoring: data analytics and intelligent platforms. By some accounts, the value derived from connected devices and sensors accounts for just a third of the total opportunity, with the remaining two-thirds derived from the platforms and analytics.
Today’s data analysis methods are capable of processing vast quantities of information in real or near-real time and returning actionable insights, forming the basis of new business services. For example, these methods can develop predictive analytics solutions that allow companies to foresee breakdowns or the need for parts replacement. This results not only in reduced costly downtime for user operations but also in lower service costs for suppliers, as routine services checks can be eliminated.
To capitalize on these benefits, suppliers have developed intelligent, broadly based platforms that can capture information and either automate the next step in the process or alert relevant personnel. For example, the system can notify the nearest engineer to the machine of predicted service issues, giving comprehensive details about the location, machine, spare parts required, and how to fix the problem. With such platforms, companies are also able to push software updates automatically and make iterative amendments to processes for optimization.
Previously, the focus in manufacturing was factory automation—typically, automating manual work processes. The IoT takes this idea much further; it uses advances in software and data analytics to allow companies to make faster, smarter decisions and pass these benefits onto their customers.
Thus, the IoT becomes the essential enabler of the smart manufacturing vision. This vision, also known as an intelligent factory or smart factory, offers a new way of organizing manufacturing processes in which different parts—from suppliers to logistics to the entire lifecycle of the product and material—become closely and intelligently connected within the corporate boundaries. The factory becomes a system of systems glued together by the data gathered in the various stages, integrated in different moments, and used for reaching the key strategic objectives of the business.
All this is created by the convergence of operational technology (OT) and information technology (IT). Historically, these have been developed as two separate domains: OT operates in real (or near-real) time, providing process control physically close to machines, while IT operates on a batch-update basis at a central location. The IoT brings these two domains together, enriching both in the process.
The IT domain gains detailed current information on machine performance and use, which can then be shared widely throughout the enterprise. OT gains data analytics used for diagnostic (why something happened), predictive (what will happen), and even prescriptive (how to prevent it) purposes. In this way, the IoT promises to optimize business processes, enable better decision making, and shorten project times by uniting these two key domains.
Exploring the Internet of Manufacturing
To identify ways in which the IoT may enhance activities in the manufacturing sector, Beecham Research divided the manufacturing sector into six key subsectors:
- Heavy Equipment
Activities within each subsector tend to use similar manufacturing processes and are therefore more likely to use similar IoT solutions. These subsectors form the basis for the inner section of the sector chart (figure 1). The second-outermost section of the chart shows typical products for each subsector; the outermost section shows IoT solutions that may provide benefits to those subsectors.
Heavy Equipment Optimizes Systems
This article further explores the heavy equipment and precision/high-tech subsectors.
Whether making trucks and trains or machinery and elevators, the heavy equipment process is quite similar—i.e., it requires multiple, additive levels and assembling fabricated materials. This sector is also a distinct manufacturing process, with many end-of-life products broken down for parts or dismantled for destruction.
Assembly lines are a key characteristic of this kind of manufacturing. It is along those lines that the IoT can have a strong impact through optimization of asset utilization, employee productivity, and horizontal integration.
The use of monitoring sensors on every component of the line will ensure full functionality of the line and predict necessary maintenance and operational changes. Head-mounted displays, supported with workflow management software, will enable a more focused interaction between the production line and worker. Additionally, wearable alarm systems and systems of surveillance cameras with video analytics and embedded intelligence can ensure security for the employees and in the entire plant.
Heavy-equipment manufacturing lines will also see increasing use of robotics (i.e., intelligent and connected devices) in place of or in collaboration with humans. Companies can take advantage of the tirelessness and repeatability of robots in combination with the intelligence, creativity, and flexibility of the human, leading to more efficient manufacturing processes. Thus, robotics will optimize production and safety.
The data that flows from all of the sensors and devices described above can subsequently be integrated with existent IT systems. The combination of machine-generated data and enterprise IT data can enable an unprecedented, horizontal view of the company from the production plant to distribution of the final product.
Data analytics serves as an essential tool for operational assessment. Predictive analytics is particularly relevant here. For example, an elevator manufacturer reduced elevator wait times by up to 50 percent by predicting elevator demand patterns, calculating the fastest time to a destination floor, and assigning specific elevators to riders to move passengers quickly and efficiently.
Intelligence-embedded devices that optimize the manufacturing processes will converge with robotics and autonomous systems. The IoT will transform heavy equipment manufacturing into a system of systems that is automated and integrated with other parts of a company’s operations. All of this will run on horizontal software platforms, also known as industrial internet platforms. The vision is one of almost autonomous production enabled by the effortless presence of intelligent devices and human creativity and flexibility.
The use of augmented reality and virtual reality will move into areas like rapid prototyping techniques for testing products and processes, and for educating employees in new manufacturing techniques.
High Tech Relies on High Speed
The methods used in manufacturing precision and high-tech goods are predominantly discrete, but they also include some process manufacturing techniques, for example, in the forging of precision instruments. Some elements are mass produced, while others are made in batches. They differ from heavy equipment in that there are typically fewer assembly levels.
The industry is characterized by the need for almost-zero error tolerance and fast-moving research and development cycles. Through the use of reprogrammable, intelligent machines and robots, the IoT can help manufacturers work to the necessary level of accuracy, while keeping pace with changing designs and assembly methods.
Computer-aided manufacturing systems will integrate with not only assembly machines but also research and development systems, allowing on-the-fly alterations. In turn, adjustments are tested using automated test equipment at crucial stages to allow for ongoing, valuable feedback. The process development execution system integrates data, legal, business, information, and knowledge processes and is optimized through the increasing data input generated by the IoT.
The demands for data processing in everyday operations and research and development in this industry are higher than in most others. Consequently, there is also a need for high-speed computing and specialist industrial control systems, such as the experimental physics and industrial control system used in some laboratories. Previously, these technologies were considered too costly to implement for many businesses. However, due to its greater connectivity and access to cloud computing, the IoT will help many companies transform critical elements of their business.
Key Technologies for Smart Manufacturing
Under the IoT umbrella, a set of technologies will strongly impact the move toward smart manufacturing:
- Data management and analytics solutions form an essential building block.
- Security is the essential feature of smart manufacturing; without it, the IoT will never develop its full potential.
- Smart manufacturing is a connectivity-agnostic environment where different connectivity solutions have a role, from cellular to satellite and from fixed- to short-range wireless.
- Smart manufacturing highlights the need for convergence between robotics and the IoT.
- The smart manufacturing vision can be realized efficiently using a horizontal management layer, known as an industrial IoT platform, that governs all the IoT connected devices and machines in the manufacturing process.
- There are a set of emerging technologies that will be increasingly relevant for the manufacturing process, from augmented and virtual reality to 3D printing.
The first two technologies in the above list deserve further attention.
Integrating Data Management with Analytics Solutions
Collecting large amounts of data from all operations must be matched by effective data processing to formulate sound decisions. The more granular the data, the more precise the understanding of what is happening on the manufacturing floor.
Today’s manufacturing plants use multiple smart systems that bring together different data sets and aim to deliver predictive and prescriptive insights. However, communication streams are still broadly focused on an individual machine basis, with information feeding into a simple reporting system that, at most, offers alerts at prescribed trigger points. The IoT offers a greater dimension to this process. It brings together vast sets of data from all machines and processes, offering deeper insight into how certain decisions made at one point may affect other connected systems.
The objective is to optimize the decision-making process in near-real time in ways that were not previously possible. Here the datasets are derived from sensors embedded in all areas of manufacturing operations, as described above.
Data management technologies now enable the handling of any data, structured or not, from any number of sources. While early machine-to-machine (M2M) implementations had to make the data from devices match a proprietary format, IoT systems incorporate internet protocol and layers of hardware, software, and networks with technologies from different vendors.
IoT systems also combine sensor-collected data with publicly available and reference data from governments and other sources, such as satellite imagery. Large global companies will have valuable historical data to make use of and differentiate on but will need the tools and data scientists to unlock that information.
Therefore, the different sensor and data types that will be collected will reside in different silos; an additional step must be taken to connect these diverse data sources into a standard format that can be analyzed. Notably, these data volumes will be large, which could result in processing delays. Considerable IT system design skills, together with subject matter expertise, will be needed to build a unifying IoT system of systems for a manufacturing installation.
It is worth noting that data ownership and sovereignty are expected to become the subject of increasing management attention. At its simplest, the entity that collects the data owns it. Once this data is shared between partners and contractors, however, the matter of who owns what and who has right of access becomes more complicated. Such agreements need careful consideration. Nevertheless, once handled, there are valuable opportunities for data owners to cross-sell to other partners in the supply chain.
Building in Security
Network security threats are increasing. While best viewed as noise that needs to be filtered out, security requirements have become an essential part of the specification for IoT solutions. IoT security must be included in solutions from the ground up; it cannot be added retrospectively.
To gain real value from the IoT, systems must be capable of interoperating. However, once anything connects to the internet, it runs the risk of being hacked, corrupted, or leveraged as part of a denial-of-service attack. Hence, it will be even more important to ensure data quality and integrity and establish well-architected security structures to mitigate increasing risks.
Manufacturing is highly competitive and of great value, particularly across subsectors such as pharmaceuticals and chipset manufacturers, so the potential for significant disruption if systems are breached is real and will inspire investment in advanced security systems.
According to Beecham Research, security in the IoT is significantly more complex than in existing M2M applications or traditional enterprise networks. The IoT ecosystem is composed of layers involving platforms, networks, sensors, systems, and users, as well as the multiple vectors of threats from external and internal actors. Consequently, significant effort is required in the identification, authentication, and authorization of all these components, as well as their users.
Mr. Duke-Woolley is an internationally recognized thought leader in the IoT market.