Autonomous Transformation is comprised of five key technologies: artificial intelligence, the Internet of Things, digital twins/simulations, robotics, and virtual and augmented reality.

The first four have each passed their inflection points, demonstrated by organizations around the world setting their strategies and investing in building or buying these capabilities. The inflection points for virtual and augmented reality, on the other hand, will arrive when the right degree of technological advancements and policy enable commoditization in both enterprise and consumer markets. They are included in the scope of Autonomous Transformation because they will become a critical lens through which developments in artificial intelligence, the Internet of Things, digital twins/simulations, and robotics can be observed, collaborated on, managed, and operated.

As with all new or reimagined technologies, the technologies of the era of Autonomous Transformation are best defined through examples. The following are three vignettes, with increasing complexity, to highlight the interplay and possibilities of these technologies.

I provide the caveat up front that the lines between Digital and Autonomous paradigms, as well as between Reformation and Transformation, are blurred, and that I am using Autonomous to describe the accumulated capability of several technologies that could also be leveraged to transition from the analog to digital paradigm without creating autonomy.

Autonomous Transformation Example 1: Product Development

The design, development, building, and scaling of a physical product is a complex process with many steps and organizations involved. Even a product as simple as a lamp still requires design, proof‐of‐concept, testing, regulatory approval, and suppliers for each component, the simplest of which contains nine components.

An analog version of the design process starts with sketches and drawings, including tables and figures of specifications such as width, diameter, length, voltage, and many more. This design would then be reviewed by manufacturers to determine the feasibility of manufacture and which components and materials would need to be purchased from suppliers, followed by subsequent research to determine suppliers and to design logistics, cost assessments to determine profitability of the product, patent applications, and submission to regulatory authorities for approval to sell within targeted jurisdictions—to name a few of the necessary steps. It is a lengthy, complex process.

Digital Reformation and Transformation have increased the speed of reviews and materially changed the design process through the addition of digital design capabilities. Digital advancements have enabled designers all over the world to collaborate in real time, and have significantly reduced the effort required to research suppliers and to model logistics. Digital capabilities have also created greater insights into past performance across verticals to inform cost/benefit analyses and have simplified the execution of cost modeling and testing scenarios.

Autonomous Transformation takes an additional step forward. The design can be built as a first‐principles (“true‐to‐physics”) simulation (i.e., a digital twin of the proposed product), which serves as a bridge between the physical and digital paradigms. This can then be tested for feasibility of manufacture based on digital twins of the organization's factories and machines. If, for example, the product called for an eight‐foot metal arm bent at a degree that was not feasible for any of the machines in the factories, this hindrance could be caught without the need for human review. Generative artificial intelligence could then be used to propose design considerations that could achieve a similar design aesthetic while achieving feasibility based on existing and planned manufacturing capabilities.

In terms of cost modeling, the Internet of Things in the factories, consisting of cameras and sensors, would have collected troves of information on quality defects, the patterns of which could be analyzed against the new design to find correlations and propose design considerations that could lower the likelihood of defects and therefore waste. This would improve the profitability calculation. The Internet of Things data could also include information on downtime, changeover, and efficiency in the development of products, which, paired with digital twins of the machines that would be used to make the product, would create more realistic cost and yield estimates than were previously possible. A digital twin of the existing supply chain and logistics could be leveraged for examining new suppliers and identifying options for fulfillment from existing suppliers, such as increasing an order for a specific type of steel from a known supplier. If it was determined that a new supplier would be necessary, artificial intelligence could recommend suppliers based on logistical, geopolitical, cost, quality, and ethical considerations and subsequently model the logistical path.

The accumulated digital representation of the entire product design developed through this process could then be submitted to regulatory authorities who could perform physics‐based simulated tests on the design to narrow the scope of necessary physical tests, increasing the product's speed to market.

The digital model developed through this process would continue to add value in future scenarios, such as monitoring the quality of supplier materials and components. Artificial intelligence could comb through defect data supplied by sensors and cameras from the Internet of Things as well as from product returns, recalls, or sensors embedded in the product to find correlations between components supplied and deficiencies or strengths in quality. Another application of the model after the design and approval phase would be in a scenario such as the need to reduce cost by 2% that was posed in a previous chapter. In this case, artificial intelligence could be leveraged to propose design considerations to reduce cost, taking into account aesthetics, quality, technical specifications, and digital twins of the factories. Using aluminum instead of steel, for example, might have appeared to be worth consideration to a human designer in the analog process, but would be eliminated from consideration in moments by artificial intelligence due to the temperatures required to harden aluminum to the technical specifications, which might not be possible with the organization's machine capabilities. This is an example of two key principles of Autonomous Transformation in action: the harnessing of human expertise across disciplines through technology, and augmenting the human creator to increase the ability to execute on existing workloads and extend capabilities into new workloads.

The remaining Autonomous Transformation capabilities that play into this process are robotics and virtual and augmented reality. Augmented and virtual reality play a similar role in this context, in visualizing the product for new designers, visualizing and supporting creative changes through the design process, and visualizing failure points within the context of the completed product for manufacturing engineers and operators. The state of robotics within the Digital Transformation paradigm has begun the shift from human engineers spending hours calibrating a machine to perform a task to using artificial intelligence capabilities to allow the robot to experiment and learn safely in a digital environment, thereby reducing the human effort required as well as time. In the Autonomous Transformation paradigm, this progresses into programming human expertise through machine teaching, which provides curricula within which the machine can train. This has been demonstrated to reduce the amount of experimentation and self‐learning required by the machine by 45 times the undirected reinforcement learning approach, achieving significant cost reductions and unlocking use cases previously unavailable to machines due to the required nuance.¹

Autonomous Transformation Example 2: Global Logistics

At any given moment, millions of goods are being moved from one location to another by car, van, truck, railroad, ship, or airplane. These include everything from chemicals, metals, batteries, and food to clothing and Star Wars collectibles purchased on eBay. A single bottle of hand sanitizer required the harvesting and shipping of chemicals to a manufacturer, which then manufactured, bottled, boxed, and shipped the hand sanitizer to stores all over the world.

Going into the analog world of global logistics may be too far back, as that included horses and telegraph communications and, later, fax machines. Digital Transformation's largest impact on logistics is communication. The current state of progress, broadly speaking, is that a material or good (down to the SKU, or stock keeping unit) can be tracked live anywhere around the world. This means that a retailer can see that the shipment of their bestselling sweater that they were hoping to restock is actually going to be approximately three days late and is on a freight sailing across the Pacific Ocean, and they should end the flash sale earlier than planned so as not to run out of stock.

A reimagining of global logistics in the era of Autonomous Transformation starts with a digital model of the planet, or at least of every logistical path. What is currently directional and reliant on crowdsourced updates for directions and estimated travel times for consumers is not reliable for enterprises because it is not possible to crowdsource updates about fallen timber over railway tracks in a remote mountain. The Internet of Things creates the possibility for sensor data to be tied to the digital twin of the logistical path of each individual item. Visibility into areas of the logistical route that are unmonitored by the cameras or sensors could adjust calculations, estimates, and pricing in real time. Partnerships with airlines, the use of drones, and purchases of satellite imagery for the areas where camera monitoring is not possible would create a method of ensuring that landslides, fallen trees, and other detectable issues are caught and addressed before they become emergencies.

Artificial intelligence can be leveraged to perform simulations of various groupings of items to determine the fastest fulfillment path and prioritizing premium shipping and guaranteed arrival dates.

When disaster does strike, artificial intelligence could determine the quickest and most cost‐efficient path for each item to get to its destination and reroute the packages and assignments accordingly.

Robotics are already being used in warehousing contexts, but as mentioned earlier, they will now be able to address previously unaddressable use cases—in this case, the picking and packaging of goods. Drones will be able to explore physical blockers with the potential to develop capabilities to remove blockers, such as finding a tree and requesting a fleet of drones that are able to lift and move the tree without the requirement of human intervention. As many have likely seen in the news, innovations are also being explored to leverage drones for delivering individual packages to consumers. Merck is a great example of this, as they have developed capabilities for delivering medication to those for whom a disruption in medication would be dangerous when a natural disaster disrupts the preexisting supply chain.²

Virtual reality can play a critical role in connecting supply chain analysts and decision‐makers with the real context of disruptions, effectively transporting them into the situation to understand the full context. This will reduce organizational friction by creating proximity, and a deeper understanding will enable more effective and creative collaboration.

Autonomous Transformation Example 3: Health Care

Autonomous Transformation empowers consumers and the organizations that serve them to increase proximity through digital means. In health care, Digital Transformation has transformed physical records to digital records. Notes and summaries from appointments, surgeries, or treatments can be accessed online, and prescriptions can be managed and requested online. Doctors can also now access records across hospitals, and the advent of telehealth appointments means that checkups and nonurgent issues can be addressed from the comfort of a patient's home.

Health care through an Autonomous Transformation lens, to expand on the example at the beginning of this book, would start with creating a digital twin of a patient's body. All known allergies, surgical records, hospitalizations, diagnoses, genetic history, medications, and symptom records would create a rich digital twin upon which analyses could be performed. The questions that patients currently answer each time they see a new specialist or doctor, or the time spent by the health care provider in reviewing previous doctors’ notes, could be significantly reduced. Previously intangible nuances such as pain tolerance could be captured as baselines for pain medications can be established and dynamically remodeled across visits, surgeries, and hospitalizations. Artificial intelligence layered on top of this digital model could interact with the doctor, who could posit a hypothetical diagnosis or treatment to the machine for feedback. If, for example, the doctor were considering whether the patient had an iron deficiency due to his symptoms, she could input that hypothesis for feedback, and the machine could share recent blood work results, cross‐check for iron deficiencies in the patient's familial records, and order bloodwork if there was insufficient data to be able to draw a conclusion. These accumulated digital models could be reviewed by doctors and patients in family planning to ensure that the family and health care team are monitoring for likely complications based on family history. For example, if a pregnant patient and her grandmother and father all suffered from complications stemming from a rare genetic deviation through the toddler years and into early youth, it would be likely that her children would be faced with the same complications. If she did not remember the exact issue she faced, or perhaps misunderstood it to be due to a different cause (which is not at all uncommon), the condition could go unmonitored and unmitigated in her child. The burden of remembering medical details of family histories can be shifted from nonmedical individuals to digital models accessible to a patient's care team.

This technological shift would also reduce the pressure on doctors to ask all the right questions during patient visits, with the potential to transform the process of reviewing documentation before and after visits to be more focused and efficient. During overnight stays, the digital model of patient pain threshold, personal baselines for vital signs, comfort levels, and allergies, for example, would augment nursing capabilities to assist with reducing the load on nurses, especially in the context of staffing shortages.

Proposed interventions such as prescriptions or surgeries could first be run in a science‐based simulation of the patient to predict likelihood of success and generate insights of potential strengths and weaknesses. A readout of this accumulated model and analyses could also be shared with insurance companies to justify chosen medical interventions.

The Internet of Things in this context could contribute significantly to the dynamism of the model. Adding heart rate, movement, calories, and exercise information would indicate the individual's level of wellness, and could create opportunities for patients to be proactively notified of unusual heart activity. This is already in the market with fitness trackers and watches, but could take a step further through integration with an accumulated digital model of an individual's broader wellness, medications, history, and genetics to generate alerts and insights more personalized for the individual that then could be shared with a medical team if the individual chose to pursue medical intervention.

Virtual and augmented reality, in this context, can create the ability for medical professionals to collaborate in a new fashion, viewing three‐dimensional models of a patient's organs, scans, joints, or fractures. A plan of reasoning and action could then be documented digitally, the output of which could be personalized to each medical professional based on their disciplines and personal preferences. One surgeon might want to review a three‐dimensional model and simulation prior to performing surgery. Another might want to use augmented reality during the surgery to keep track of vitals, the plan, and to account for any anomalies.

From a robotics perspective, as robotics are increasingly used in various medical interventions, digital models, combined with artificial intelligence (specifically, deep reinforcement learning, for the technologists reading this), could serve as the input to a machine task. This, paired with using machine teaching to capture human nuance in the medical context, could greatly augment medical professionals and assist with the higher degree of demand than available supply of medical professionals and expertise.

This example is much more uncomfortable for the average reader and for technologists to discuss, because health care strikes to the most fundamental level of existence: life itself. Beyond personal comfort levels with the subject, there is the added complexity of a heavy degree of regulation (and for good reason), as well as privacy concerns.

This illustrates another key element of the era of Autonomous Transformation. The technological capabilities that comprise this era can only be harnessed in the context of new overarching systems: of the health care system, the health insurance system, the educational system, health care policies, and the broader ecosystem supporting health care. Transformations of this nature are critical and inevitable given the advancements of technology and society. Leaders have the opportunity, at the onset of this new era, to design the transformation to achieve a more human future.

A Note on Blockchain

Blockchain is an important technological advancement, and a critical enabler for the next iteration of the decentralized World Wide Web (Web 3.0). Although it is not a core technological component of Autonomous Transformation, it can be leveraged in combination with Autonomous Transformation technologies to engender trust within organizations, between organizations, and between organizations and consumers.

Within the context of the global logistics example discussed earlier, blockchain could be leveraged to provide visibility to the chain of custody between organizations and to the consumer. A layer deeper than logistics is in the actual sourcing of materials, goods, and labor involved. This is one of the key arenas in which explorations and investments in blockchain are taking place, with the aim of creating and maintaining certifiably ethical sourcing.

The health care example almost necessitates a discussion about blockchain. Currently, individual health care records are distributed across systems within every health care provider a patient has visited. Each hospital, insurance provider, pharmacy, and medical laboratory has at least one record stored in a database, whether on premises or in the cloud. This redundancy renders maintaining current records across all systems practically impossible, and, more importantly, decreases both the privacy and security of important personal data.

Decentralizing this information would remove redundancies and provide patients with full visibility to their medical records. At the point at which a provider needed to access medical information in order to examine or treat a patient, the patient could leverage their private key, review the access level requested, and grant the provider access to the blockchain for a designated period of time, which the patient could revoke at any time. Any new information added by the health care professional would be signed by his or her personal key, which would mean that patients would not have the ability to alter information without the approval and participation of the medical professional who entered that information. This would be important for maintaining the integrity of medical records.

Blockchain, much like the technologies within the scope of Autonomous Transformation, has tremendous capability to advance progress and contribute to the creation of a more human future, but in order to achieve its potential, it will require systemic design, the creation of Profitable Good market dynamics, and surprising and remarkable partnerships.