A number of this year’s predictions are based on analytics, with big data, data collection and visualization, and predictive analytics noted as top gamechangers. Read on to learn the role these technologies and more will play in the manufacturing industry this year.
***William Surphlis is the managing partner for Grant Thornton, Productivity Improvement, a dynamic team that leads the charge in business analytics and management consulting.
The future of manufacturing, as many other industries, lies in business intelligence and data analytics. Never before have we had so much information at our disposal, and the ability to effectively decipher it to make our businesses better. That’s why, I believe in 2017, the most essential manufacturing technologies will be those which allow manufacturers to harness data and leverage it to improve processes from all angles — driving new efficiencies and record results in such areas as quality control, employee morale, customer delivery and production, to name just a few.
1. Data collection
All manufacturers should have a reliable and cohesive way to collect data in real-time if they hope to thrive in 2017 and beyond. With a host of data collection and monitoring solutions on the market, there’s definitely no shortage of options. The key is to find the one that works best with your existing environment and systems — and make sure your time and resources are spent on collecting data that is relevant to the improvement of your operations. While challenges still remain in collecting information from legacy systems, many software companies are starting to lean towards more open tools, networks and communication protocols to make the dream of a fully-connected plant floor a little more plausible.
2. Machine learning
Machine learning — a type of artificial intelligence (AI) — is expected to be a big business in 2017 and beyond, and it has enormous potential in the manufacturing sector. Whether you’re using machine-learning algorithms to improve quality control or keep the supply chain running efficiently, there are plenty of opportunities to be had. These algorithms learn continually and are capable of finding optimized outcomes quickly. This allows smart machines to go beyond basic inspection tasks and identify difficult-to-detect defects — or mitigate supply chain issues (such as a parts delay) thus avoiding line shutdowns.
3. Big data analytics
While the collection and visibility of data is one thing, transforming it into usable information that can improve operational capabilities is quite another. That’s why it is essential to centralize data gleaned from machine learning solutions — as well as production reporting tools, supply chain analytics, distributed storage and distributed processing — so all information is readily available in one place. It also may help to invest in the services of a business analyst to transform the numbers into information that can be easily utilized by management.
4. Data visualization
With meaningful business information in hand, forward-thinking manufacturers are finding more innovative ways to share that information with the rest of the company as a means of improving productivity, and this trend will likely continue in 2017. Taking data visualization to the next level by enhancing digital dashboards and displaying televisions or video screens throughout the plant floor is one way to do this. By presenting real-time metrics in a prominent position for all to see, organizations have the potential to improve such things as productivity, employee morale and engagement, on-time delivery, customer service and plant safety.
5. Data storage
Regardless of how much data you have, and how well it’s collected and analyzed, it won’t be worth much if you can’t access it and use it in a timely manner. That’s why the Cloud works. Not only does it make it easier (and faster) to share information company-wide, but unlike brick-and-mortar data centres, the Cloud is flexible, allowing you to increase your storage as your needs change.
Craig Resnick is the vice president of ARC Advisory Group. He covers automation supplier and financial clients, with 30 years of experience in marketing, business development, and strategic planning. He graduated Northeastern University with an MBA and BS in Electrical Engineering.
1. Advanced analytics, AI and machine learning become IIoT enablers
The use of analytics tools has been the business intelligence (BI) platform, supplemented by enterprise manufacturing intelligence (EMI). These tools excelled at helping users discover and understand the underlying reasons and details about what happened and why. Now, with the industrial space becoming much more dynamic, manufacturers are turning to advanced analytics, AI and machine learning to support predictive and prescriptive analytic solutions. By connecting previously stranded data from smart sensors, equipment and other assets with advanced applications and predictive analytics in the Cloud, the Industrial Internet of Things (IIoT) is becoming a strategic enabler to improve manufacturing performance.
2. More industrial network devices are living on the “edge”
The industrial network “edge” is being populated today by devices such as: Ethernet, wireless and cellular gateways; Ethernet switches and routers; and wireless access points (WAPs). Traditionally relied upon to bridge information technology (IT) and operations technology (OT) environments or bring legacy sensors, devices and assets into automation or enterprise architectures, today’s network edge products target sensor-to-Cloud integration to further industrial Internet-based strategies designed to improve business performance. The edge is where the assets and the associated data to be analyzed by enterprise applications reside, ranging from protocol conversion gateways for interfacing disparate networks to sensor- to-Cloud integration and edge computing.
3. Your assets have a digital twin
A digital twin is the asset’s virtual representation, an archive of historical and real-time data, drawings, models, bills of material, engineering and dimensional analysis, manufacturing data, and operational history that can be used as a baseline when benchmarking performance. Similarly, real-time data acquired via integrated sensors or external sources is used for analytic tasks, including condition monitoring, failure diagnostics, prescriptive and predictive analytics. Knowledge gained adds value to asset life, improving efficiency, reducing downtime, anticipating failures, and for continuous improvement at the design and manufacturing levels. With a digital twin, closed-loop design now extends through the entire product lifecycle.
4. Using simulators for training, leveraging augmented and virtual reality (AR/VR)
As manufacturers hire new employees, they are implementing training that uses simulators to convey plant knowledge, leveraging technologies such as gaming, augmented/virtual reality, and 3D immersive, with wearable devices, such as the Microsoft HoloLens. This enables being able to replicate real plant and job functions, controls and assets, providing a high-fidelity experience. Simulation improves learning and is effective in developing skills to deal with unanticipated plant situations, thus increasing workers’ confidence in performing their job functions and ability to deal with an emergency. Other applications of simulation include testing and validate new software, performing system migrations, and program testing and validation.
5. Acceptance of disruptive technologies: SaaS, virtualization, big data, convergence, and BYOD
Manufacturing functions, such as material and energy procurement, product quality, and production management, are performed through software-as-a service (SaaS) provided by a third-party. Virtualization technologies reduce computing hardware, software and IT support costs along with energy consumption. Manufacturing leverages big data to convert data into actionable business intelligence. IT/OT convergence enables integrated families of business applications that utilize both real-time and transactional data. “Bring Your Own Device” (BYOD) enables operators, supervisors and managers monitoring plant and factory performance to use their own mobile devices. Social networks create virtual user groups within and between plants, as well as to the technology suppliers.
Andrew Hughes is a principal analyst at LNS Research with his primary focus being research and analysis in the Manufacturing Operations Management (MOM) practice. He has 30 years of experience in manufacturing IT, software research, sales and management across a broad spectrum of manufacturing industries.
This year, we have taken a broad look at what is happening in manufacturing and present here our top five predictions. There is a sixth that is impossible to predict but something we think is likely in the coming 12 to 18 months — an IT/OT mega-merger. We expect the acquisition spree for smaller analytics companies to continue and there might also be a big takeover of an IT company by one of the operationally focused giants. We shall see.
1. Predictive analytics specialists flourish
We predict non-traditional vendors displace traditional domain specific applications in predictive analytics with 40 per cent of new deployments coming from the emerging solutions providers by year-end 2017. When we look at the rise of the IIoT over the last two years, most of the application focus has been on Asset Performance Management (APM), which can be continuously improved through better predictive analytics. It is therefore no surprise to see many specialist start-ups developing analytics solutions for APM. Other traditional areas where new analytics can already deliver valuable results include process optimization and proactive quality. New analytics capabilities are being applied to help process engineers find new answers to questions like “How do I avoid process degradation?”
2. Big commitments to IIoT platforms
IIoT platforms are being offered (or promised) by an ever-increasing band of large automation and software vendors. We predict that, before the end of 2017, at least 2 to 3 Fortune 500 manufacturing companies will make a corporate commitment to a specific IIoT platform. It is this type of commitment that IIoT vendors have been waiting for. Today, there are few applications running on these platforms; once the platform is sold, the opportunities to develop applications for, and port them to, the platform will accelerate quickly.
3. Multimarket monolithic MOM is dead
Last year we predicted the demise of monolithic MOM and the rise of modular solutions that can run on IIoT platforms as well as stand-alone on premise. We have been pleasantly surprised at the speed of change to more modular solutions, even from the bigger and more traditional players. As manufacturers move towards IIoT platforms delivered by the big OT players, MOM solutions will inexorably shift to a set of apps that provide flexible and easy to implement solutions. However, the MOM suppliers who focus on a few specialized markets undoubtedly have longer to make the transition because their selling points are configurability and focus on special requirements. Today, their fitness for purpose is still more important than their flexibility.
4. Traditional automation companies self-disrupt
The ISA-95 model describing the control hierarchy has defined the scope of delivery of automation suppliers for years. From ‘sensor to controller through MOM to ERP’ has been a great way of defining the business and products delivered by the large automation suppliers. In 2017, this comfort will change and automation suppliers will launch products and deliver solutions that will disrupt the status quo. We have already discussed a new MOM model and can expect to see more integration of devices directly to IIoT gateways developed by the control companies. These will bypass levels 3 and 4 of the ISA-95 model. In addition, Cloud-based analytics will interact with, and directly control, plant-level devices. The breakdown of the ISA-95 hierarchy will offer opportunities to third-party software vendors but the real wins will come when the automation suppliers choose to self-disrupt — that way they will avoid self-destruction.
5. IIoT platforms will attract developers, fast
As industrial analytics started to move from traditional to IoT-focused, there was an early warning from the industry that there would be a critical shortage of data scientists. To date this has not occurred. However, with the interest in IIoT platforms and the number of offerings being marketed, we predict 2017 will be the year of the IIoT coder and data scientist — developers will hit critical mass in more than one platform and in multiple analytics packages. The IIoT platform companies that have not set up the infrastructure for the modern young coder to have access to their platform will be also run by the end of 2017. On top of that, we expect several high-end value-added resellers of traditional control solutions to make big moves into IIoT delivery.
Sivakumar Narayanaswamy is the research manager, Industrial Automation & Process Control, at Frost & Sullivan.
1. Data-driven manufacturing
Data is the new currency, as data and analytics form an integral part of decision making in the manufacturing sector. Adoption of data-driven manufacturing (big data) delivers significant value as it improves asset performance by decreasing unscheduled downtime, leading to increased productivity and cost reduction. The influx of large volumes of structured and unstructured data, owing to the increase in connected assets, will push the establishment of a scalable framework, such as the Hadoop cluster, to manage, integrate and process big data. In 2017, the industrial data analytics market is expected to witness a spike in demand for predictive and prescriptive analytical platforms.
2. Industrial robotics
Coupled with big data and analytics, use of robots on the shop floor is enabling exploitation of machine data in order to deliver data-driven insights and decisions that aid in mapping productivity and assessing asset maintenance. Our research finds industrial robots are undergoing a phase of favourable transformation characterized by new players, as well as mergers and acquisitions spearheaded by IIoT opportunities that are expected to yield revenue of $46.19 billion in 2017. Articulated robots and collaborative robots will drive robotics in 2017. New business models include collaboration-as-a-service, plug-and-play models and robotics-as-a-service.
3. Industrial Cloud
With the Cloud emerging as an innovation platform for manufacturing, Frost & Sullivan predicts Cloud-based technology will transform from being offered as a product to being offered as a manufacturing service, though this will be a long-term transition. Besides enabling seamless interactions between devices and services connected to the Internet, the industrial Cloud will evolve into a technology that enables device connectivity that paves the way for edge data analytics. Security concerns, insularity, knowledge of Cloud complexities and uninterrupted interwork availability are some of the challenges that must be overcome for higher adoption of Cloud services. We expect more automation as IT vendors venture into this market, either directly or by forging partnerships with niche Cloud service providers.
4. 3D printing
As the manufacturing world moves from mass production to mass customization, 3D printing is expected to enable the development of an agile manufacturing environment, resulting in the reduction of lead times — from conception to the production — by 70 per cent or more. Wide-scale adoption of polymer 3D printing in the consumer electronics and metal prototyping automotive industries are major drivers in the growth of 3D printing. As products are made-to-order in real time, there will be a reduction in logistics costs or a decrease in warehouse space, resulting in a complete overhaul of the conventional supply chain. While automotive, aerospace and medical are the top three industry adopters, process industries are likely to start gradually adopting this technology as well. 3D printing as a service is poised to replace traditional manufacturing business models with the emergence of integrators providing services across the value chain driving the mainstream adoption of 3D printers.
5. Industrial mobility
Several industrial verticals, like automotive manufacturing and food and beverage, have been demanding access to critical manufacturing information, anytime and anywhere. Acceleration of BYOD can be related to the growing acceptance of wireless technology on the plant floor. The need for remotely operated assets is also driving the need for building a tablet application instead of a conventional control panel or HMI device. Being able to view multiple assets’ health, production schedules and reports from a smart phone or tablet has been enabled by the superior computing power now available and the communication networks that provide real-time access to information. However, higher adoption of industrial mobility in the manufacturing floor is impeded by the vulnerability faced to “secure” the critical production information. Progress in securing the communication channel by employing newer versatile technologies is likely to bolster the adoption of industrial mobility in 2017 and beyond.
This feature was originally published in the January/February 2017 issue of Manufacturing AUTOMATION.