Industrial digital transformation is the term applied to the demanding journey that an organization undertakes when it integrates business model change, process improvement and cultural shift, often leveraging digital and emerging technologies to create and exploit new opportunities in the marketplace.
In this overview, we identify and discuss the main drivers of transformation, and consider what transformation means for organizations in terms of its value and benefits.
We will refer to industrial digital transformation in the context of companies from commercial or public sectors which deal with physical assets, factories and field operations, in relation to business-to-business scenarios and the transformation involved in making improvements to such products, equipment and operations. When the transformation involves software or business enhancements for asset light or pure tertiary services companies, including business-to-consumer scenarios, then we refer just to digital transformation. When we mean both, we refer simply to transformation. Our use of the term industrial includes both commercial and public sector.
Exploring industrial digital transformation
Industrial digital transformation may often involve a radical rethinking in the use of technology, culture, people and processes in the enterprise. This can lead to a fundamental change in business performance and outcomes, as well as in how customers perceive the company. Figure 1 is an easy way to look at the transformation process. It shows that culture and technology changes go hand in hand with the business process and business model changes.
The technologies used to help drive industrial digital transformation may include one or more of:
· Internet of Things (IoT)
· Cloud and Edge Computing
· Artificial Intelligence (AI) and Machine Learning (ML)
· Big Data & Analytics
· Blockchain
· Robotics
· Drones
· 3-D Printing
· Augmented Reality (AR) & Virtual Reality (VR)
· Robotic Process Automation (RPA)
· Mobile Technologies
New technologies continue to emerge, so this list is not meant to be exhaustive. The main goal of such transformations is to gain competitive advantage, drive new revenues, improve productivity and efficiency, as well as enhance customer and stakeholder engagement. The term technology or digital technology, in the context of industrial digital transformation, is not limited to software or information technology (IT). It may also include physical, chemical or biological/life sciences and related technologies. For example, in the context of autonomous vehicles it can be LIDAR (Light Detection and Ranging) or a more efficient car battery. In the industrial safety context, it can be a sensor or a system for fall detection or a thermal scanning camera for infectious disease detection or prevention.
According to the Customer Insights & Analysis group of the International Data Corporation (IDC), worldwide investment in industrial digital transformation-related initiatives is expected to exceed $6 trillion over the next four years (2020–2024) (see https://www.businesswire.com/news/home/20190424005113/en/Businesses-Spend-1.2-Trillion-Digital-Transformation-Year). Smart manufacturing will account for a large part of this spending. Other sectors, like finance, retail and logistics management and transportation, will also undergo large-scale industrial digital transformation.
In April 2020, at the height of the Covid pandemic, while reporting quarterly earnings, Microsoft CEO Satya Nadella said, ‘We’ve seen two years’ worth of digital transformation in two months’. With a changing global landscape the pressure to innovate has only increased, and the pace of digital transformation seems unlikely to lessen in the coming years.
Identifying business drivers for industrial digital transformation
The power of transformation applies to some or every aspect of an organization. It can generate business value, agility and resilience. The importance of resilience is shown at a time of local or global crisis.
The different forces that help to shape the industrial digital transformation in an organization are shown in Figure 2. An industrial digital transformation often entails a series of big bets or bold steps, to achieve large-scale benefits or competitive advantage. This differentiates transformation from regular generational changes, which are often linear or comprise a series of small, gradual steps. Humans have climbed mountains through the ages and eventually climbed Mount Everest. However, an analogous series of incremental improvements would not have landed a human on the moon. This is probably an extreme example of scaling ‘new heights’ in the history of humanity. But so is a Level 5 autonomous car (see https://www.nhtsa.gov/technology-innovation/automated-vehicles-safety#topic-road-self-driving) compared to the first car with an internal combustion engine (built in 1885).
Figure 2 shows the change agents on the left-hand side such as business process and model changes, with support from technological and cultural shifts. These are often forced by traditional competitors or by disrupters. Regulatory changes, and the expectations of customers as well as shareholders, change over time. Transformation helps to ensure that productivity, profitability and social responsibility improve and align with the interests of stakeholders.
Business drivers in the commercial sector
In the commercial sector, the need for industrial digital transformation is often driven by two alternative kinds of strategy:
· Defensive strategies
· Offensive strategies
Defensive transformation strategies aim to protect the business from competitors and disrupters. Most car manufacturers started manufacturing electric vehicles as a defensive strategy. According to Moody’s, traditional US car manufacturers lose $7000 to $10,000 per electric vehicle. The major reason the car manufacturers continue to invest in electric vehicles is because this market is expected to grow by almost 20% in the next decade. With breakthrough innovations expected in battery and related technologies, the cost of production is expected to go down.
While most automobile manufacturers have pursued a defensive strategy, Tesla exemplifies the adoption of an offensive strategy, where it is trying to disrupt the rest of the industry. A large proportion of both outlook and forecasts in the automobile industry today have been driven by Tesla, which started in 2003 and is newer than most US and global auto giants. Today, it reduces some of its losses by charging a price premium, differentiating itself through becoming a status symbol and offering driver-assist technologies. Tesla is not a profitable company as of early 2020, but is aggressively reducing losses because of the lifestyle status and innovation that enables it to charge a premium price. Tesla is a good example of industrial digital transformation at work in the automotive industry, while the Tesla Semi is targeting disruption of the trucking industry next.
Tesla’s approach to future-proofing involves equipping its cars with the hardware necessary to make the cars increasingly autonomous in near future via over-the-air (OTA) updates. This will increase Tesla’s market valuation as well as the value of Tesla cars for their current owners. While Tesla, being a newer company, is free of cumbersome legacy processes, there are still areas where Tesla can transform itself. The company’s reliability score, especially in relation to the Model X, has been poor, mainly due to challenges in the design of the door. In Tesla we find the paradox of high emotional attachment, for it is a fun car to drive, alongside not-so-great quality scores. (See: https://www.forbes.com/sites/petercohan/2020/07/25/the-tesla-paradox-highest-emotional-attachment-lowest-quality-says-jd-power/#3bd0e5a97594.)
The transformation at Tesla is effectively implemented across its entire value chain, with the integration of its products, services and operations. Tesla is an example of a connected car, which allows the creation of digital twins of its cars. A digital twin, by one definition, is a virtual representation of a physical object or a system that can be used to improve the performance and efficiency of the physical counterpart. Tesla uses the digital twin of the car to provide new services with OTA updates to the software. Related to the digital twin concept is the idea of a digital thread, which is often used in the industrial manufacturing sector to improve the product quality and the throughput across the entire lifecycle of the product.
In the next section, we will look the drivers for transformation in the public sector, where the concept of profitability has a different meaning to the meaning it has in the private sector.
Business drivers in the public sector
Government Digital Services leapt into public consciousness in the USA in 2013 with the implementation of the Affordable Care Act (ACA), also known as Obamacare. For a variety of reasons, the development of the federal healthcare exchange, that is the front-end websites and the back-end databases and processes known as HealthCare.gov, started late and failed miserably. See: https://www.gao.gov/assets/670/668834.pdf.
Health and Human Services (HHS) had used the same process that the government has used for many years to develop and deliver solutions and had achieved roughly the same results that government technology projects had achieved for decades. A team within an agency within HHS developed a set of requirements, published an RFP, accepted bids, selected a vendor and then waited for delivery of a product. That turned out not to meet requirements, and, in fact, failed to deliver the capabilities required for the successful launch of the new healthcare marketplace.
Faced with the failure of the Administration’s signature legislation, the Obama administration did something different than past administrations and projects leaders. They put out a call to the private sector for help. A group of engineers led by Mikey Dickerson worked round the clock for months to repair and modernize HealthCare.gov. In a moment of clarity that comes all too infrequently, members of the team and others within government recognized that HealthCare.gov was but one example of a larger problem with the way that the public sector builds and buys technology solutions. See: https://money.cnn.com/2017/01/17/technology/us-digital-service-mikey-dickerson/index.html.
Many of the leaders of the HealthCare.gov rescue effort, including Dickerson, became the core of the United States Digital Services (USDS), part of the Executive Office of the President reporting to then US Chief Technology Officer, Todd Park. The USDS, 18F at GSA and other digital services teams were created due to a general recognition within the federal government that technology projects took too long, cost too much, failed too often and, even when considered successful, rarely met the needs of the public they were designed to serve.
The US government spent close to $75.6 billion on various IT projects in 2014 (see https://myit-2014.itdashboard.gov/), across all federal agencies, including the Department of Defense, large cabinet-level departments , such as the Departments of Labor, Transportation and Agriculture, medium sized agencies such as the Environmental Protection Agency, and small agencies including the Small Business Administration and the Nuclear Regulatory Commission. In addition, according to the Standish Group, of the over 3,000 IT projects with labor costs that exceeded $10 million that the government executed between 2003 and 2012, only 6.4% were considered successful. Over 41% were complete failures, meaning that they had to be scrapped and restarted. This problem was not limited to the federal government or the development of new solutions: it plagued state and local governments too.
Government inefficiencies are often blamed for these failures, but the true cause is both more complex and more understandable. It is simply impossible to specify all the functionality of a large software system and anticipate all complexities before the development process has begun. Nor is it reasonable to expect that, when the technology lifecycle is frequently less than two years, the technical design and needs of users can be fully specified years in advance. Simply put, it was clear after decades of large-scale failures that the longstanding practice of spending years gathering requirements, months or years selecting a vendor, and then years waiting for a big-bang delivery of a solution that had been developed in seclusion wasn’t working (and possibly never had worked).
In addition to the fact that the traditional government project development process didn’t deliver consistently working software that met the requirements initially specified by the project team, the traditional model rarely delivered software that met the needs of end users. Government software was frequently developed with a small number of stakeholders in mind. Stakeholders in the government are generally the individuals who sponsor or fund a solution, but they rarely represent the bulk of users of a system. For example, if the design of a new timecard system was stakeholder centered, it would be designed to streamline the process for the handful of individuals in accounting who manage the back-office processes. In contrast, a user-centered design would focus on the needs of the large majority of users who interact with the system occasionally.
Another example of the power of user-centered design is the US Environmental Protection Agency’s (EPA) eManifest system. This is a voluntary, fee-based system successfully deployed by the EPA in 2018, but it didn’t always seem certain that the project would be successful. In 2015, as the project was floundering and under increasing scrutiny, the EPA’s new Chief Technology Officer, Greg Godbout, led a relaunch of the program. One of the first things he learned was that the project team had never talked to a single end-user. He arranged for the team to go on a listening tour. During this tour the project team learned that the solution they were proposing to develop wasn’t what the users needed. They were trying to solve the wrong problem. Talking to users before writing code allowed the project team to reset early when the costs were low, rather than after the project had been completed and when the cost of changing course would have been a complete reboot costing millions of dollars. The key idea of user-centered design means that the transition to digital services moves the government closer to the public, allowing the government to develop solutions that more closely match the needs of its constituents.
Traditional development processes don’t just hamper the delivery of new capabilities to the public, they put existing service delivery at risk. The COVID-19 crisis has exposed this vulnerability to the millions of out-of-work Americans who were unable to file for unemployment benefits in a timely manner, as states were unable to scale up the systems that process unemployment claims, due to antiquated architectures and reliance on obsolete hardware. During the early days of the state lockdowns, individuals attempting to file claims reported system crashes, unavailable websites and hours-long hold times or busy signals as many state systems required that individuals called to complete their claims that had been started online. Many of these systems reside on mainframes and are written in obsolete languages such as COBOL and do not follow the best practices used in coding today. Commonly they feature messy ‘spaghetti’ code, written to preserve then-precious processing power, at a time when commenting in code was non-existent.
Note
The outbreak of COVID-19 tested the limits of many older government IT systems and highlighted the need for modernization of legacy systems. Many of these legacy systems were written in COBOL, a language that hasn’t been taught at most universities since the 1970s. You can read here about how keeping these systems running has created a need for COBOL programmers, much like the Y2K bug did in the late 1990s: https://nymag.com/intelligencer/2020/04/what-is-cobol-what-does-it-have-to-do-with-the-coronavirus.html
In addition to the need to implement technology to support government policy, to deliver new capabilities to the public and ensure the reliability of existing services, crises such as 9/11, the COVID-19 pandemic, hurricanes, earthquakes and wildfires all demonstrate the need for government to be fast and nimble. COVID-19 requires the combination of fixed sensors and contact tracing applications to control the spread. Governments must be able to expand the capabilities of existing systems and deploy new, previously unanticipated solutions in order to respond to crises. Ironically, these same crises demonstrate that the government can be nimble. With a state of emergency declared and procurement rules suspended, one federal agency hired contractors and redeployed a government loan program application over a weekend while a state agency engaged a firm to provide call center software, a ticketing system and agents to augment their unemployment system over another weekend. With a sense of urgency and without the constraints of a highly regimented procurement system, governments can move fast and serve constituents better.
As the case became clear and heroic successes such as Healthcare.gov were demonstrated, individuals and teams at all levels of government around the world began to explore institutionalizing digital services across government.
In the next section we look at how emerging technologies are becoming an integral part of the transformation journey.
Technology drivers for transformation
No specific technology is a solution for all areas of industry, but rather each technology must be paired with an appropriate problem statement alongside an understanding of its limitations. In addition, several technologies are in what one may consider the ‘hype’ phase of their maturity cycle and it remains to be seen if they will be viable in the future. A practitioner is well served by taking an objective stance in relation to these technologies rather than climbing onto the hype bandwagon.
A specific example relevant today is that of blockchain. Blockchain was conceived as a solution for anonymous untrusting parties to transact with each other and avoid the double spending problem (see Satoshi Nakamoto, Bitcoin: A Peer-to-Peer Electronic Cash System, 2008. Available at https://bitcoin.org/en/bitcoin-paper.) Blockchain can be viewed as a solution looking to solve a problem in the industrial setting. To be shown to be a viable solution the problem must satisfy the core premise underlying Bitcoin — namely transactions across anonymous untrusting parties. In addition, a large number of industrial use cases involve rates of transactions that are much more suited to traditional databases than a distributed ledger.
The Hype Cycle for Blockchain Technologies (July 2019) from Gartner shows a 5- to 10-year timeframe before blockchain becomes mainstream and has a transformational impact across industries (see https://www.gartner.com/en/newsroom/press-releases/2019-09-12-gartner-2019-hype-cycle-for-blockchain-business-shows). We advise vetting the blockchain application closely to ensure it is the best fit in any particular scenario. In the Hype Cycle, from an industrial perspective most of the top contenders — like use of Blockchain in Transportation and Logistics, Blockchain in Supply Chain, Smart Contracts and Blockchain in Insurance — highlight the likelihood of a blockchain-led transformation in the supply chain and distribution space within the next decade. However, the use case of the company Colu, around the digital currencies for cities, encouraging them to spend their money locally, has not been as successful as one might have hoped. (See https://www.wired.com/story/whats-blockchain-good-for-not-much/ for more details.)
Some technologies are of course much more mature than blockchain. Figure 3 gives a preview of some of the other technologies having potential as drivers of transformation:
We will now walk through some of the components illustrated in the figure.
· Sensing: Before one can talk about digitization it is imperative to ensure one has a solid foundation for sensing and collecting data across the enterprise. Data collection can scale from sensors in the process flow — potentially augmented by IoT devices at the edge which collect and aggregate data to external factors related to logistics and demand sensing. Another falling under the heading of sensing are machine vision systems and their associated algorithms, which analyze image data via edge computing and send summarized data to data aggregation systems. Such sensor data has to be analyzed in the broader context of enterprise data which may originate in enterprise resource planning (ERP) and other manufacturing or maintenance systems.
· Data aggregation: Rather than keeping data in silos, one must aggregate it in a common location — which can range from an internal data lake hosted on premises to a Storage as a Service (STaaS) facility hosted by an off-premises cloud services provider. Typically, the latter also offer additional services to entice customers to move to their platforms. Data aggregation is critical to enable everyone involved in the enterprise to get a single version of truth. Connectivity and integration across various segments of the enterprise are key to enable this capability.
· Analytics: This comprises of a suite of methodologies that operate on the aggregated data, some of which are listed below.
· Statistical analysis: This is fundamental for all smart manufacturing efforts. Initial use cases revolved around statistical process control (SPC), first proposed by Walter Shewhart in 1924 with the invention of the control chart (see: Walter Shewhart, ‘Economic Control of Quality of Manufactured Product’, American Society for Quality Control, 1931). This found widespread use during World War II, leading to the Six Sigma methodology. Shewhart greatly influenced W.E. Deming, who created the now-famous funnel experiment, resulting in a greater reliance on statistical process control that initially hindered the application of modern control systems methodologies to industrial problems. In 1951 Box and Wilson introduced response surface methodology which led to the development of designed experiments. This was the first attempt to systematically develop input — output models of industrial processes in order to drive the process to the optimal operating point. In addition, statistical analysis is widely used in inventory management across the industrial supply chain.
· Artificial Intelligence (AI): This is a broad field covering traditional rule-based systems, statistical machine learning, and recently deep learning. Figure 4 shows the relationships between these various methods:
Some examples of specific techniques that have been popularized across these fields are rule based systems and fuzzy logic, inferencing under traditional Artificial Intelligence, tree-based classifiers, support vector machines, and convolutional neural networks for image processing and deep reinforcement learning.
One must take care to select the correct methodology for the task. As is often the case, the simplest methodology generally leads to the most robust and sustainable solution.
· Optimization and simulation: Optimization and simulation are critical tools for implementing any kind of decision system. Such systems can function either in an automated mode, for example scheduling systems, or can be used to guide a human to make decisions by simulating and optimizing various scenarios (i.e. the user asks ‘what if xyz were to occur’ and the system will simulate that condition and optimize system performance to give the answer).
· Visualization and dashboards: As data moves through the analytic engines there is still a need to visualize it from time to time. For a person to make sense of the data, there is a need for the analytics to distill the raw information across all sources into a few key metrics that will be meaningful to users.
As AI applications proliferate, the user will need to get less and less involved in mundane decision making, needing only to respond in situations where an autonomous decision system is unable to make a decision, or to course-correct an erroneous decision. As such the metrics should reflect not only the overall health of the industrial system (be it a manufacturing plant, or the entire supply chain) but also relevant metrics to track the robustness of the AI models.
One also hears the term big data being associated with machine learning. There is definitely overlap between the two, but big data focuses especially on the data storage infrastructure, on the platform for executing analysis of this data in an efficient manner, and on computationally scalable algorithms for feature extraction (or dimensionality reduction) in order to make large volumes of data amenable to machine learning algorithms.
One of the most successful big data platforms is Hadoop, with its distributed file system and the capability to efficiently implement MapReduce algorithms that provide highly scalable processing of large volumes of data. (See: Dean, J. and Ghemawat, S. (2004), MapReduce: Simplified Data Processing on Large Clusters, Communications of the ACM, 51(01): 137–150)
While these digital technologies are adopted it is important to keep the security and safety of people and property in mind. As the physical world gets connected to the network due to the Internet of Things (IoT), cybersecurity considerations become of paramount importance. When digital twins are created and stored, they could be the targets of security breaches that seek to obtain access to restricted data or operating details that are otherwise not public information. The 9/11 hijackers had reportedly trained on flight simulators and software games (see: https://publicintegrity.org/national-security/authorities-question-criteria-for-access-to-flight-simulators/). It is vital to ensure that the digital twins of power plants or nuclear plants and other critical infrastructure do not fall into the wrong hands.
Why industrial digital transformation matters to business
The internet and web applications, and the easy availability and low cost of massive amounts of computing power and storage, have revolutionized the way that businesses operate. In the process they have reset the competitive landscape. In some cases, industrial digital transformation is a competitive advantage, but in others it is simply the minimum effort required to stay in business. For many organizations, digital transformation is a do-or-die proposition.
Industrial digital transformation can serve one or more of three purposes for business:
· Improve internal processes, reducing cost and increasing competitiveness
· Streamline delivery of existing solution within an existing business model to reduce cost or improve customer service
· Transform a business completely, resulting in new products and business models
A true digital transformation is a disruptive innovation that fundamentally changes the user experience. This new experience, if delivered properly, will delight the customer and provide the business with insights into how to better serve that customer in the future. It can also enhance customer support processes, leading to lower support costs and new insights about the customers.
Industrial digital transformation is not simply the automation of existing processes using new technology, but rather the reengineering of existing processes and products to deliver fundamentally different solutions. A simple example of an internal process improvement is the routing of a document for review. When a document was routed on paper, it would move to each individual reviewer in sequence. Once that document was digitized, it could continue to route to each reviewer sequentially. However, if the process were redesigned it might be routed to all the reviewers except the final approver concurrently, shaving days or weeks off the review process.
At the product level, industrial digital transformation allows for the creation of entirely new products that could not exist before digital solutions existed, disrupting entire markets. For example, the ridesharing applications Lyft and Uber would not exist if not for the digital disruption of business models. Before the advent of the smartphone and sophisticated algorithms to rapidly match riders and drivers and to manage pricing, to keep supply and demand evenly matched, these car sharing services could not have existed. They have disrupted both the taxi and car rental markets.
Digital transformation matters to businesses because virtually all businesses are being disrupted. New entrants are arriving with lower costs and new approaches to the existing business or with new business models that cannibalize their business. Incumbents must transform their culture, processes and technologies to compete and thrive in this changing landscape.
Quantifying the business outcomes and shareholder value
The decision process in large public or private organization is often driven by strategic goals or the value of investments to stakeholders, while also aiming to make the organization stronger and more sustainable. As a result, any new initiatives beyond incremental efforts to preserve the business, go through business case of return on investment (RoI) analysis. Hence it is important to understand the key benefits of industrial digital transformation to the business. The desired outcomes of the digital transformations are often:
· New digital revenues
· Productivity gains
· Corporate social responsibility
Let’s consider these outcomes qualitatively here:
New digital revenues
In this scenario, transformation is used to drive new lines of business or new digital revenues for the existing business. A good example is servitization of the product, whereby the company tries to wrap a physical product with services that bring in recurring revenues; for example, buying the scheduled maintenance service when buying the car. This prevents the service revenue from going to the after-market parts and third-party service providers. More complex examples include a jet engine provider selling ‘thrust by the hour’ for the aircraft or, the ‘power by the hour’ model. To build the business case for this type of outcome of industrial digital transformation, the proposed investment is weighted against the possible new revenues. (See: https://knowledge.wharton.upenn.edu/article/power-by-the-hour-can-paying-only-for-performance-redefine-how-products-are-sold-and-serviced/)
Productivity gains
In this scenario, the primary goal of industrial digital transformation is to improve the bottom line and drive efficiency. Take the example of a wind turbine owner or operator. The cost of servicing a certain type of wind turbine, including the oil change and servicing the turbine bearings, is about $8000 per service. To prevent excessively frequent servicing, which would result in higher routine maintenance costs and not servicing when due, which could lead to expensive damage to the wind turbine, the company decided to move to condition-based maintenance (CBM). They added sensors to the oil to record viscosity levels and other particulate levels. This allowed the company to come up with the optimal servicing scheduling through remote monitoring — a great example of realizing productivity gains by means of industrial digital transformation.
Social responsibility
Often both private sector and public sector companies look at transformative ways to fulfill their corporate citizenship goals. The business case for these may include both tangible and intangible benefits. For instance, an airline may set stringent goals for carbon offset and look for transformative changes to accomplish that.
In the case of autonomous cars, it is important to think through the breadth of the impact. As autonomous cars become mainstream, they will have an impact on how roads, traffic signs and even cities and airports are designed. Likewise, they may have a profound economic impact not only on the auto industry but also on utility providers due to electric requirements, and on employment in the automobile and the trucking industry.
Finally, auto insurance and (in the USA) the Department of Motor Vehicles would also have to adapt to the changes wrought by autonomous cars. Hence, a technology-led digital transformation of the automobile has profound social-economic and political impacts. Change management and a phased approach must apply not only to ‘technical’ aspects of the transformation but also to change of the business landscape and the societal impacts.
To identify opportunities for industrial digital transformation, much can be gained by examining the normal business cycle that an industry goes through to manufacture goods or to provide a service. Consider a new product introduction, the typical stages of which are illustrated in Figure 5.
Industrial digitization can play a crucial role in each of these stages. To look briefly at some examples:
· Concept: This is the initial ideation stage that helps define the requirements for a new product. Digitization can help here by providing machine learning solutions that combine unstructured data to spot key customer trends. Several suppliers offer platforms for analyzing social media messages to gauge positive or negative sentiment related to features in existing products.
In addition, given the lead times required to move from concept to production ramp it is helpful to be able to forecast the product feature sets possible when the product is in general availability, as well as expected sales volumes. Machine learning and data mining can provide significant benefits in this area.
· Design: Digitization can help in the design process by allowing greater collaboration between designers. Collaborative tools that allow designers across the globe to work together on a common platform — even being able to share and edit drawings concurrently — can greatly speed up the design process. In addition, digitization provides the means to reuse components already in use at the company, can limit later headaches around raw material SKU management, and can help drive the economies of scale that keep costs down.
· Prototype and validation: Rapid prototyping is key to evaluating the design for fitness and functionality, and to make any final revisions before the product is released to manufacturing. Additive manufacturing can play a key role in rapid prototyping of mechanical parts. For electronics there are special companies that specialize in small batch orders with quick turnaround to get samples back to the customer quickly. These companies leverage computer-integrated manufacturing to quickly reconfigure tooling between customer orders.
· Customer trials and compliance and regulatory testing: Being able to send prototype samples to customers, the manufacturer can get rapid feedback on new product features. As Steve Jobs once said: ‘People do not know what they want until, you show it to them’ (see Isaacson, W. Steve Jobs: The Exclusive Biography. New York: Simon & Schuster, 2011). Rapid prototyping provides an avenue for doing so. Customer trials using prototypes can be sped up by employing technologies such as digital twins. Using these to conduct tests under extreme environmental conditions may help guide the design towards meeting regulatory requirements.
· Manufacturing: This encompasses several areas, each with its own sets of challenges and opportunities for digitization. Manufacturing comprises of not just the factory or network of factories but the entire supply chain network. This area is teeming with digitization opportunities.
It is not hard to find publications and videos relating to the use of augmented reality, control rooms, machine learning/AI, detailed real-time simulation models (see for example demonstrations from GE on models of their aircraft engines — also referred to as a ‘digital twin’) and autonomous planning and scheduling.
This is perhaps because manufacturing and processes involved are relatively well understood, and one can readily control the sensors and metrologies in these. This contrasts with, for example, applying natural language processing and sentiment analysis to unstructured data for determining new features which may entice customers during the concept phase. Connected products and operations provide opportunities to improve customer support operations and drive efficiencies.
Lastly one should keep in mind that the aim of digital transformation is to enable the following:
· Faster time to market with a cheaper cost per unit
· Managing and reducing environmental footprint
· Reducing risk to production by enabling digitization of the supply chain and the workforce.
The aim of a commercial enterprise is to maximize profit, revenue, and market share, and digitization technologies implemented correctly provide opportunities for visibility, efficiency, and agility.
How will industrial digital transformation impact the future of work? This will be a key concern from the perspectives of those responsible for driving the strategy as well as execution of the transformation. The growth of automation and use of ubiquitous AI will profoundly change how we work. The ‘lights out’ data center is one good example of that. Likewise, ‘cobots’, or collaborative robots, where humans work alongside the robots in factory settings, is another pointer to the future of work.
The explosion in use of remote conferencing technologies like Zoom and Webex in the first quarter of 2020, during the Covid-19 crisis, is another example of the changing nature of work possible in extreme scenarios. Tele-medicine grew as well in this period, and the regulatory landscape was relaxed (see https://www.hhs.gov/hipaa/for-professionals/special-topics/emergency-preparedness/notification-enforcement-discretion-telehealth/index.html).
The ‘gig economy’ or the shared economy has been possible due to transformations in related industries. As we move to autonomous vehicles such as autonomous trucks, will it disrupt the profession of truck drivers? Interestingly, during the first quarter of 2020, long haul truck drivers were in very high demand for the distribution of food and groceries to the retail industry.
Conclusion
In this introductory overview you have learnt what industrial digital transformation is, and have seen how cultural, technological and business drivers are involved. Its benefits for organizations include increased competitiveness and efficiency, and it may bring about new products and services for customers. The quantifiable benefits can include new digital revenue streams, productivity gains, and/or new possibilities for meeting social responsibility goals. Opportunities for industrial digital transformation abound along the entire product creation process, from conceptualization to manufacturing and delivery. Indeed they are to be found throughout an organization’s value proposition. Identifying them, however, is more than a matter merely of understanding what technologies might be relevant to a particular set of organizational circumstances. Rather, or in addition, it requires thinking strategically and holistically about all aspects of an organization’s operating context, its ultimate aims, and — by no means least — the goals and values of its stakeholders and customers.
Further reading
George Westerman and Didier Bonnet: Leading Digital: Turning Technology into Business Transformation, Harvard Business Review Press (2014).
To purchase the book on Amazon, click on- https://packt.live/3p0nKDy
