The digital transformation of manufacturing isn’t easy. But the right selection of technologies coupled with smart implementation strategies can enable manufacturers to realize the benefits of the Fourth Industrial Revolution. By Anant Kadiyala
Throughout the history of manufacturing, companies have sought to leverage the power of emerging technologies to improve what they make and how they make it. Over the years and epochs of industrial change, manufacturers have sought to streamline production, reduce errors, decrease waste, and improve safety, among other improvements.
However, this era of Industry 4.0 may not follow the same technology curve as the previous three generations. Companies that are ahead of the curve are innovating on new business models, new service models, new production techniques, better inventory management and sustainable sourcing.
On the flip side, companies that are slower to embrace change often face faster business erosion of customer trust and loyalty. The rate of turnover in the Fortune 500 is at an all-time high. In the year 2000, for example, Apple was not on the list of the top 100 revenue earners. Now it is third on that list. Industry analysts and experts attribute such churn to the rise of the tech sector. The famous proclamation of software is eating the world still holds true, and continues to devour all industry verticals, including manufacturing. This tech-infused dynamic is changing the pace and nature of business cycles. But according to a survey by the World Economic Forum, only 14% of the executives are ready for Industry 4.0.
However, it doesn’t have to be that way. With the right use of technologies and adoption affordances, manufacturers can cross the digital chasm and enjoy the rewards. We have a new generation of technologies that forward-thinking manufacturers are harnessing to shorten product cycles, deliver higher product quality, establish tighter customer feedback loops, improve safety, offer flexible product buying options, enable meaningful customer experiences, and also to foster rapid innovation and iteration.
Technologies such as cloud, IoT, AI/ML, blockchain, additive manufacturing, cognitive robotics, 5G, and AR/VR have been evolving and maturing rapidly over the past few years. While they all have bespoke origins, they have been gaining adoption in the B2B world (5G is still fairly nascent, so the jury is still out). Each one of these technologies is valuable to manufacturers and businesses in their own right, so companies have been gradually employing them for business advantages in recent times. However, as is with most technology breakthroughs, when the technologies are mature enough, and can also be leveraged in conjunction with each other, the cumulative benefit is often non-trivial and transformative.
But change is easier said than done. Even companies that are eager to embrace change often stumble with digital transformation initiatives. With these initiatives, there is a friction between having to serve the cash cow versus innovating for the new world to be. In addition, lack of available skillsets, miscalculating risk-reward and opportunity costs, a poor selection of projects, and the incorrect application of technology are often the cited reasons for these mishaps.
However, there is a way to avoid these mistakes. Before embarking on big hairy audacious goals (BHAG) it is essential to fathom the gestalt of business, people, and implementation complexities. Companies must find prudent ways to harness them in ways that are beneficial in the short term, and yet pave the way for a brighter future. Let’s now explore how these transformational technologies can be applied effectively, while achieving these vital goals.
The mushrooming of so many varieties of pragmatic applications over the past 10 years would not have been possible without cloud architecture.
Getting Beyond Simple IoT
Manufacturers have been adopting Industrial Internet of Things (IIoT) technologies to streamline operations. Most of the current implementations focus on basic instrumentation of equipment and data visualization, which is a good place to start. IIoT unlocks greater value when insights from sensor data can be utilized to optimize business processes. For example, we have sophisticated algorithms to perform asset monitoring and predict if a machine component is likely to fail in the near future. In most vendor stacks, this prediction ends up as an alert on the operator’s dashboard or an email. Thereafter it becomes a manual process.
While this may be better than having no instrumentation or prediction, we can certainly do better. What would be truly valuable is when incident/event predictions can help the company move from scheduled maintenance to condition-based maintenance, streamline field service operations by scheduling the right technician at the right time, and provide the technician the full context of equipment performance, incident, and diagnosis, all with just a few clicks. With the right technology stack, all of this functionality should be available out of the box. The next level of sophistication is implementing digital twins to get a holistic view of the machine lifecycle, performance, utilization, depreciation, and KPIs on a single pane of glass.
Among other IIoT use cases, streamlining production monitoring, fleet monitoring of trucks, and tracking worker safety (especially in hazardous environments) are the most common ones.
When implemented right, the operational entropy that companies accept as the cost of business can be squeezed out. That is the net benefit of these IIoT solutions. By predicting and minimizing production incidents and interruptions, throughput and quality are improved.
This kind of sophistication can sound surreal, but it is not. Advanced IIoT platforms today harness advances in machine learning and artificial intelligence (ML/AI) to deliver intelligent solutions that are more flexible and scalable than traditional rules-based systems. In fact, the moniker Artificial Intelligence of Things (AIoT) is on the rise as a result. Instead of using physics-based analysis, which tends to be brittle, these systems leverage software-based intelligence to manage the physical world.
For example, Machine Halo offers computer vision-based monitoring of equipment and facilities. 3D Signals is pioneering with sound and vibration analysis algorithms to detect equipment performance anomalies. In warehouses and factory floors, such non-invasive solutions are gaining favor to detect certain kinds of conditions and anomalies. But while image and vibration analysis are helpful, they are not the best fit for all situations. For other use cases, traditional heat, light, viscosity, or pressure sensors would be required. As discussed above, an incident detection from these systems could be a source of input for condition-based maintenance and efficient field service.
Companies are also leveraging other forms of machine learning in areas like customer service. Due to the maturity of Natural Language Processing (NLP), a specialized area of AI, digital assistants (sometimes referred to as virtual assistants) are now available. Companies are providing self-service mobile apps in the form of digital assistants for a range of queries around product, support, service, and sales. These apps can interpret plain English text, convert that to a technical query for backend systems such as ERP or CRM and then respond back to the customer in plain English. Voice assistants like Apple Siri work in similar fashion, except that they do the additional step of speech-to-text (and vice versa) conversion.
Cloud as an Enabler
As we have seen so far, ML/AI-based solutions are viable and helpful for many kinds of specific tasks. This mushrooming of so many varieties of pragmatic applications over the past 10 years would not have been possible without the enabling cloud architecture. Even though academia has been conducting research in ML/AI algorithms since the 1960s, the needle really started moving with the maturity of cloud computing over the past ten years.
Cloud technology enables seamless elastic scale up and scale down of an enterprise application infrastructure. This allows developers to quickly deploy massive computing at miniscule cost compared to the fixed capital costs of an on-premise data center for the duration needed to compute the algorithms. This elastic flexibility and cost effectiveness led to a rapid improvement in the sophistication of ML/AI algorithms. As ML/AI became more accurate, IIoT became smarter, viable, and more useful for streamlining a variety of operations.
In addition, cloud computing is also the key reason for the rise of software-as-a-service (SaaS) applications. Most modern B2B software applications are built in the SaaS model, as they allow vendors to deliver features faster and enable better security for their customers. As the pace of business increases across all industry verticals, running on-premise solutions is becoming the Achilles heel, and so companies are moving from on-premise to cloud to stay nimble and adapt quickly. Cloud economics make it a better value proposition for cost-management and cybersecurity risk mitigation as well, compared to most on-premise solutions.
Another technology that is complementary to IIoT is augmented reality (AR). For scenarios such as visualization of machine data, equipment repair, and next best action, AR is pragmatic. AR overlays digital data and images on the physical world. This gives us a rich interactive experience with inanimate objects. Imagine a scenario where a complex machine needs urgent maintenance. In a tight labor market, the right skillset is often hard to find in a timely manner. In such scenarios companies are leveraging AR to visualize IoT data and to offer step-by-step repair instructions to the technician. Just like a GPS navigates us with turn-by-turn directions, AR glasses can guide the technician.
Most AR platforms allow for use of voice commands (for hands-free experience) and remote co-browsing (dial-an-expert) to facilitate quick and effective first-time fix rates for the machines. In warehouses, AR is making headway to guide stock pickers with next-best-action recommendations, and to optimize their routes. Virtual reality (VR) is a similar visualization technology that is finding its way for use cases such as remote site navigation and worker training. While both AR and VR are making headway, the headsets and their battery life need more improvements before we will see widespread deployments of them in the field. While most of today’s deployments are limited scale pilots, these are technologies that offer significant business value through human error reduction and greater employee productivity.
Additive printing and cognitive robotics are changing the nature of production, product design, and product inventory. Autonomous vehicle maker Local Motors 3D prints 80% of the vehicle body. Sneaker manufacturers like Nike and Adidas have already released 3D printed shoes to the market. While much of additive printing is polymer based, improvements in laser technology are showing promising signs for printing of metal parts as well. As this technology matures and unit economics improve, the nature of manufacturing for intricate low-volume components is bound to change.>/p>
“As ML/AI became more accurate, IIoT became smarter, viable, and more useful for streamlining a variety of operations.”
The Rise of Blockchain
All the technologies we have discussed so far are deployed within the confines of a company. However, blockchain offers a unique technical architecture that opens up new frontiers for efficient and transparent cross-company collaboration, coordination, and streamlined workflows. Circulor leverages blockchain to track conflict minerals like tantalum. Certified Origins, a producer of extra virgin organic olive oil, leverages blockchain to track and trace its products. CargoSmart has seen a 30% increase in operational efficiency of shipping documentation with blockchain. Several spare parts makers and high tech manufacturers are exploring blockchain to track and trace spare parts and machine component provenance. In addition, blockchain is also used to simplify and optimize supply chain finance, contract/warranty management, and recall batch/lot tracing.
With blockchain, each participant of the business network owns and runs a node. Therefore, sharing and viewing of certain information across companies is simplified with the right security, privacy, and confidentiality measures, thereby allowing efficient cross-company workflow enhancements and also better business transparency.
The industry is at a critical inflection point where most manufacturing operations will be reimagined
Getting Started with Transformation
As the above examples illustrate, these emerging technologies have proven their value in the enterprise. As with all software advancements, it first happens slowly and then suddenly. In fact, many of the technologies discussed are sophisticated enough to bring tangible and immediate ROI to companies. The platforms that are SaaS based also offer a lower TCO and affordable entry points, even for small and mid-scale manufacturers. For example, Noble Plastics, a custom injection molding and design company, adopted a SaaS-based IoT platform to automate its production runs and to run a lights-off third shift.
Given so many opportunities and multiple technologies to pick from, how might a manufacturing company get started on the right foot with digital transformation? How does one manage risk, while implementing and deploying these new technologies? Here are a few guidelines:
- Pick the right problem to tackle: While there may be many potential starting points, it is most prudent to start small, and scale from there. Start with making an inventory of all potential projects/improvements. Then carve out the ones that have good business impact, high pain points, and moderate technical risk. When one applies these filters, the options quickly whittle down to a very small list. The projects that fit this criterion are often good places to start.
- Pick the right technology platform: While there are many choices in the market from established vendors to niche startups, be mindful of the complexities around integration, maintenance, upgrades, and security. Technology debt can compound very quickly with an incorrect platform choice. The ideal platform should be business user friendly, enable companies to go-live in a matter of weeks (not months or years), and have a SaaS-based architecture for its core software. The right platform should allow the manufacturer to start small and grow the implementation incrementally, so that the digital transformation initiative can scale up incrementally across multiple business divisions in a low-risk manner.
- Pick the right starter team: The success of pilot projects can make or break a well-planned initiative. Companies that sputter also end up paying opportunity costs since competition and new entrants could move in rapidly. Therefore, the right team with good fire-in-the-belly and strong team dynamics will make a successful outcome of a project more likely. For such pilots, small, effective teams often deliver more reliable wins than large teams. Once initial success is achieved and the team gains some implementation and deployment experience, scaling up to other teams and business lines gets much easier.
- Manage objections and concerns: With transformational technologies like the ones we discussed so far, some employees tend to have objections and concerns. It is important to help them understand how the digital initiatives could help them be more effective in their jobs and mitigate some of the common challenges of manual and analog methods.
As the adage goes, time and tide wait for none. Technology progression and the ensuing creative disruption, too, wait for none. Companies that embrace change, experiment with new concepts, and harness them to offer engaging products and services, streamline operations, and align partner incentives are the ones that reap the benefits of the new world.
The early 1900s are a great example of how new technologies of the day such as electricity, telephony, and mechanization changed the world for the better. Along with advancements in the fields of law and finance, they formed the modern global economy that we know today. Once again, with the new generation of technologies, we are at a similar critical inflection point where new business models, customer experience concepts, and manufacturing operations can and will be reimagined across the board. Change also brings great opportunities. With the right strategy, technology platform, and methodical execution, manufacturers can position themselves for a bright future. M