3

Workshop Two, Part One

OPENING REMARKS

Workshop series co-chair Lt. Gen. Michael Hamel (USAF, ret.), independent consultant, highlighted key topics shared by Department of the Air Force (DAF) leadership during Workshop One. He recalled Gen. David Allvin’s focus on the imperative to accelerate to establish a stronger posture against emerging global threats, as well as Gen. David Thompson’s description of the Space Force’s position as a digital force and its dependence on the Air Force for information capabilities. He noted Gen. Allvin’s interest in understanding specifically how information will improve decision making and how transformation efforts could be best sequenced to enable rapid acceleration, and Gen. Thompson’s request for feedback on whether the DAF is on the right track with its digital strategy.

Workshop Two chair Dr. Pamela Drew, former executive vice president and president of information systems, Exelis, welcomed participants to the second workshop in the series, which would reveal experiences with and best practices for digital transformations from the perspectives of information systems experts and managers in industry, academia, and other government agencies. She posited that the DAF could bring to bear these lessons learned in its journey of digital transformation.

WHY DIGITIZE?
A PANEL PRESENTATION

Mr. Robert Tross, principal, Deloitte Digital, explained that digitization should prioritize the customer with human-centered design and elevate the human experience. Digitization that embraces an agile mindset leads to rapid change and faster decision making. Because “culture eats strategy for breakfast,” he suggested that organizations who are beginning a transformation journey adopt the Scaled Agile Framework^®

(SAFe^®)¹—a delivery model that removes tribal thinking, aligns teams, helps move to fast programmatic decision making, and fosters stronger organizations. With this decentralization of trust, the hierarchy shifts and decisions can be made at a lower level. He remarked that data are the “fuel” for organizations that embrace artificial intelligence (AI) for data-driven decision making and asserted that digitization is no longer optional. Millennials and tech savvy individuals expect digital platforms that enable self-service; with AI, machines can complete manual tasks, and highly skilled individuals can focus on more impactful work.

Mr. Tross described the three pillars of realizing a digital vision: (1) embrace iterative, fast, lean, high-touch, value-added, point-of-contact delivery; (2) perform demonstrations, vet technologies, and utilize software to collect real-time feedback so that customers receive useful products; and (3) strategize by focusing on outcomes instead of outputs (i.e., focus on what you want to achieve to meet a user need, not what you want to build). Having witnessed slow, bureaucratic decision-making processes in the federal government, he advocated for changing the approach to problem solving: instead of asking “what technology are we going to use,” ask “how am I going to solve this problem with X technology.” This shift from “what” to “how” enables acceleration and, ultimately, true transformation.

Mr. Tross shared three recent examples of successful transformation led by Deloitte Digital. First, the company delivered 17 applications in only 18 months, each of which solves a particular problem. Second, the company digitized a U.S. Department of Agriculture nationwide food distribution system for the Indian Tribal Organizations (ITOs) by putting the system online and moving it to the cloud. Because the ITOs were comfortable with technology and engaged in this process, Deloitte Digital also delivered an iPad app based on customer feedback to better meet user needs. Third, the company began working with the Space Force and Cisco ~10 months ago, using Salesforce to power a digital transformation for commercial satellite usage and consumption. He highlighted this final example as a solid use case for customer relationship management, because this project requires managing relationships with both providers and distributors, and ensuring that the digital transformation aligns with Gen. John Raymond’s intent.

In closing, Mr. Tross described several ongoing challenges and lessons learned in digital transformation:

• Eliminate siloes, such as the barrier between IT and business, to avoid the derailment of programs.
• Create an organic feedback loop that involves real users; simulated users are not a substitute for real users.
• Instead of changing for the sake of change, go “all-in” on agile: begin with the outcome in mind, and create roadmaps and time frames to hold people accountable.
• Have a supportive, agile contracting structure that is fixed-outcome and capacity-based instead of fixed-price.
• Embrace minimum viable products that can be placed in the hands of users quickly instead of striving for perfection.
• Hire a team that is comfortable with transformation, because digitization will feel chaotic to those new to the concept.
• Use platforms that support rapid change to experiment, fail fast, and adapt.

Ms. Danielle Ullner, partner and managing director, Boston Consulting Group (BCG) Digital Ventures, explained that her organization works with other institutions to invent, launch, and scale game-changing digital products and platforms as part of broader digital transformation efforts. She echoed Mr. Tross’s mantra to fail often, fail fast, and then pivot to something else.

The right multidisciplinary team makes it possible to transform at warp speed: in 7 years, BCG Digital Ventures has built more than 160 digital solutions.

___________________

¹ The website for SAFe is https://www.scaledagileframework.com/, accessed November 23, 2021.

She emphasized that digital transformation is no longer a choice for organizations. However, only ~30 percent of organizations achieve successful digital transformations because digital transformations are difficult to execute; delivering such fundamental change at scale in large, sometimes global, complex organizations is challenging, especially when there are short-term pressures and leaders may have to put their careers on the line. The cost of failure is high—large IT write-downs, poor user experiences, missed growth and productivity opportunities, and loss of time. She suggested six steps for organizations, which could increase the success rate of digital transformation to ~80 percent:

1. An integrated strategy with clear transformation goals;
2. Leadership commitment from the chief executive officer through middle management;
3. Deployment of high-caliber talent, including a balance of change managers and product managers focused on the end user;
4. An agile governance mindset that drives broader adoption;
5. Effective monitoring of progress toward defined outcomes; and
6. Business-led modular technology and data platform.

Ms. Ullner stressed the need for leadership to avoid “analysis paralysis,” which can destroy an activity before it can be scaled from pilot to funded program. These six key success factors are imperative to develop a “bionic organization”—in other words, instead of making one-off, standalone digital improvements, intentionally combining the human element with the technological capabilities to ensure that organizations can thrive in the face of change and uncertainty, drive breakthrough innovation, and solve significant problems (see Figure 3.1).

Ms. Ullner shared a case study on transforming a multinational utility with everyday innovation at speed and scale. Although this is a real-world industry example, she emphasized its relevance to the DAF. For example, the utility operates in a complex stakeholder environment, has constant resource issues, and is eager to gain competitive advantage. The utility’s specific challenges included an inability to meet customer expectations for seamless digital experiences, the threat of digital-first utilities leveraging technology to change industry rules, a new generation of workers expecting increased efficiency through technology, fragmented and data-poor planning efforts leading to poor capital investments, and new regulations forcing cuts to operating budgets. The utility had a bold vision to reinvent its business through digital transformation with optimized maintenance (e.g., better use of data and predictive analytics to maintain and repair equipment), dynamic grid operations, on-demand deployment of the field force, and seamless customer experience. A new chief information and digital officer worked to develop a trusted relationship between the utility’s business and technology communities, and employees were invested in delivering results in real time. As a result, the organization as a whole became more empowered. BCG Digital Ventures developed a program for this utility that was anchored to five principles: (1) start small but think big, (2) learn by building, (3) deliver impact at speed, (4) build capabilities just-in-time, and (5) change from the inside out. BCG Digital Ventures also helped the utility to create outcome-focused digital roadmaps with a portfolio of products that reimagined “business as usual” processes. She added that any transformation process is most effective when it is stage-gated and milestone-based, and when ineffective initiatives are promptly eliminated. With BCG Digital Ventures’ assistance, the utility launched four new products in less than 18 months, supporting teams with enabling talent development, governance, and technology platforms. During the next phase of the project, BCG Digital Ventures and the utility strengthened the operating system to scale capabilities and support successive waves of digital products. More than 40 new digital hires were made; staff have been exposed to digital processes, agile mindset, and quick decision making; and products are delivering value.

Ms. M. Nadia Vincent, digital transformation and innovation executive advisor, Digital Transformation Leaders, observed that because many organizations begin to digitize either in direct response to a regulation change or to regain advantage against their competitors, the transition likely reveals additional organizational problems. To avoid this outcome, it is critical to create a strategy for digital transformation prior to implementation. She presented a model that could help leaders strategize their vision for digital transformation—the Digital Wheel of Wealth (see Figure 3.2) reveals the many opportunities of the digital age as well as the six characteristics of digital intelligence:

1. Broken time, distance, volume, and physical barriers. Time is the most precious asset to manage; once time is optimized, it is possible to define and achieve more outcomes. There is less distinction between the physical world and the virtual world. Digital transformation enables easier collaboration across distances, and computers have become so powerful that organizations are increasing investments in storage.
2. New efficiency. Digitization increases productivity, accessibility (specifically with AI and mobility), measurability (when tasks are delegated to computers), visualization, understandability, and assistance in achieving goals.
3. Limitless possibilities. Progress comes in the form of creations from imagination, innovation of new processes, and improvement of existing processes.
4. Leadership. Because computers are still guided by humans, visionary leaders who are dreamers, protectors, innovators, and rainmakers are needed to inspire people at every level of an organization and to enable action.
5. Growth and expansion. Changing people’s mindsets and creating new habits requires investment in people. Expansion of organizations, communities, businesses, and societies begins with individual human growth.
6. New relationships. New technologies allow for greater connectivity. The man/machine relationship, in particular, is the most challenging. Instead of replacing humans, machines could

1. collaborate with human intelligence to produce and achieve at a higher level.

Ms. Vincent explained that before creating a digital strategy, organizations should consider the following questions: What opportunities can you foresee for your organization if barriers are broken? What digital efficiencies could benefit your organization? What are possibilities you can imagine for the future of your organization? Who are the leaders that will inspire the organization at various levels? What are possible future growth and expansion opportunities for your people and your organization? What benefits could emerge from improving relationships in your organization? Once these questions are answered, and potential opportunities and capabilities are recognized, it is possible to define a strategy for digital transformation. She emphasized that data need to be given special attention within this strategy, because an organization cannot survive without data.

In closing, Ms. Vincent suggested the development of guidelines and processes for people to adopt the transformation and make investments in the right leadership team. Because innovation should be an ongoing process, she advocated for organizations to create innovation business units.

Dr. David Bray, executive director, GeoTech Commission, Atlantic Council, and former chief information officer, Federal Communications Commission (FCC), explained that digital transformation is 20 percent technology and 80 percent “hearts and minds.”

He described his experience leading a bipartisan commission to review all research and development efforts in the intelligence community in 2011. As the fourth person to attempt this effort in only 3 months, he emphasized that the barrier to success was not one of technical execution; instead, it was the difficulty of managing the expectations of Congress, the troops, and contractors during this “disruption.” Thus, it is critical for organizations to have leaders who are aware that transformations ask people do things with which they may be unfamiliar or uncomfortable.

In a similar experience, Dr. Bray arrived at the FCC as the ninth chief information officer in 9 years. Upon arrival, he considered an important question: What is the level of acceptance for change in the organization? Many government organizations have “change fatigue”; when that is the case and people are reluctant to change, he continued, it is important to consider new strategies. He noted that the Air Force often relies on doctrine to codify people’s actions, but he wondered whether the current doctrine rewards and incentivizes the desired behaviors. Without that component, people will not change and digital transformation will not be successful.

Dr. Bray explained that the first step in digital transformation is to understand the “why,” by identifying the problem that needs to be solved. Bray emphasized that the “why” for Congress (e.g., to be more effective with available resources) may differ from the “why” for commanders and airmen (e.g., to become nimbler and more adaptive in future environments), as well as from the “why” for contractors. He pointed out that the incumbent contractor workforce is often the most resistant to change. It can be difficult to break longstanding relationships with contractors; however, in a digital transformation, often different and fewer contractors are needed.

Dr. Bray advocated for the government to embrace private sector partners on its journey to innovation, and encouraged the Department of Defense (DoD) to use the combination of cooperation and competitiveness that exists within the services to provide incentives so that change also comes from the bottom up. This approach is more effective than creating an isolated “innovation office” at the top, because digital transformation involves everyone in the organization. He suggested adopting the “melted iceberg approach,” where people share their approach to a challenge, others provide feedback and recommendations, ideas are spotlighted, and a competitive rivalry within the service is created to move ideas forward. He emphasized that digital transformation is enabled by a shift in mindset, in which everyone is considered to be a valuable part of the broader effort. Allowing people in an organization to voice their concerns is another way to shift people from being “problem holders” to “problem solvers.”

Dr. Bray underscored that a successful transformation starts with low-hanging quick wins, with bigger, bolder actions emerging 1 year into the transformation. He cautioned that even at this point, 10–15 percent of people likely will still resist or even sabotage efforts, and those people may ultimately need to be reassigned or released. An important part of leadership during a digital transformation is helping people navigate the loss of “the old way of doing things”; it is critical to consider how to blend new thinking with old thinking, providing opportunities for collaboration and for programmatic units to highlight their accomplishments. He emphasized that successful digital transformation results from changing programmatic processes as much as from changing technology; there is no separation between IT and mission, as IT is part of the mission. In closing, he suggested assembling a diversity of perspectives both inside and outside of an organization to try to manage any friction that could hinder a transformation.

Open Discussion

Dr. Drew described several themes that emerged throughout this session of the workshop. While senior executives in industry understand the need for government to be more efficient and will find ways in the long term to support the government, not everyone in an organization will share that perspective. She also reiterated that although technology is important, it is not the foundation for digital transformation. Instead, digital transformation is enabled by breaking down barriers, reimagining processes, and reinventing approaches. Because the people carrying out existing processes might resist these changes, it is crucial that they are part of the solutions. She added that eliminating funding barriers across components and departments of the Air Force is particularly difficult. Gen. James (Mike) Holmes (USAF, ret.), senior advisor, The Roosevelt Group, observed that although Air Force leadership has a transformation strategy, it does not have funding to recruit the right talent to conduct agile operations in support of that strategy. This cycle persists in part because the Air Force continues to buy “stuff” instead of focusing on investing in its network for operations.

Reflecting on the value of “quick wins” to build momentum for transformation, Col. Scott McKeever, director, Chief of Staff of the Air Force Strategic Studies Group, asked about methodologies for project selection. Ms. Ullner described four criteria for choosing the first program of a transformation: (1) high value and impact; (2) a low level of implementation difficulty (i.e., select a project with known, minimal technological and operational risk that can be readily addressed); (3) few dependencies, so as to reduce delays that could hinder success; and (4) a middle-management-level champion. Dr. Bray added that early projects should also demonstrate cost savings (e.g., moving the cost from an operation and maintenance expense to a capital expense). It would also be beneficial if an early project could demonstrate a new model for security. Ms. Vincent suggested using design-thinking to create value in different phases of a project. Mr. Tross noted that sustainability should be a consideration, even in the “quick wins” phase. For example, to determine which projects to do first, his team built rationalization scorecards based on complexity level and potential benefits (i.e., how many users a process has, how many customers it affects, etc.). This exercise instills organizational rigor, and could be used to measure success and to determine a course for future projects.

Lt. Gen. Hamel observed that large, complex organizations such as the Air Force have numerous internal processes, stakeholders, and customers to prioritize and engage. Dr. Bray suggested thinking about this challenge from the perspective of a venture capitalist. He asserted that a transformation effort should not come from a single office; it requires a communications strategy that is accessible to all stakeholders, which describes the why and how of the transformation as well as identifies appropriate behaviors. Mr. Tross added that SAFe enables the collaboration of mission and IT—and once that unity is achieved, it becomes possible to prioritize and engage. Furthermore, SAFe enables the management of a few programs around a few technologies, which is more scalable and digestible than managing a portfolio of many disparate technologies.

Lt. Gen. Hamel described a unique challenge for the Air Force: commercial elements have to be purpose-built and designed to operate across multiple operating domains. He wondered how to avoid a fixation on “stuff” but have a trustworthy infrastructure purpose-built to support missions. Dr. Bray responded with the following transportation ecosystem analogy: the transformation effort would have one team focused on “roads” and the safety features of those roads, and another team focused on assembling the features of the “car” that will drive on those roads (i.e., the application). The car should have the assurance that it can safely drive on those roads. Translating that analogy to the digital transformation, such an approach ensures that each time there is a new effort, there is no need to start from scratch: the security is baked in, no matter who hosts the platform.

LESSONS FROM OTHERS “LIKE THE AIR FORCE”: PANEL ONE

Mr. Rahul Welde, executive vice president of digital transformation, Unilever, explained that Unilever is one of the world’s largest consumer goods companies, and it is driven by a mission to make sustainable living commonplace. Unilever has a presence in more than 190 countries, and more than 2.5 billion people use its products (400 brands) every day.

He observed a commonality between the Air Force and Unilever—both are a force for good that serve the needs of people. Although digital transformation in industry is different from that in other sectors, he continued, shared core principles exist: no matter the sector, opportunities and challenges arise related to the substantial impact that technology can have on people’s lives. He noted that the recovery effort from the COVID-19 pandemic has been led by new technologies, just as the recovery effort from the 1929 Great Depression was led by new industries as part of the Industrial Revolution.

Mr. Welde remarked that digital transformation is a broad term with varied meanings. Unilever understands digital transformation as the intersection of technology, consumer behavior, and a large customer supply chain-organizational interface. Digital transformation is thus about end-to-end business and how to impact that spectrum through tools, technologies, and capabilities. He described four “thinks” to consider for digital transformation within any organization:

1. When people, processes, and technology work in balance and in tandem, there is exponential impact on transformation.
2. It is important to harness the power of data and insights about consumer needs. Unilever’s Foundry, People Data Centres, and internal studios (where it shares information about products with consumers) are unique, new capabilities that have emerged to deal with the enormous amount of available data. Technology enables real-time data processing and decision making, which has implications for the fundamental capabilities of any organization. Agility is needed to meet the challenges of the supply chain and evolving consumer practices: speed (with precision) is the new currency.
3. Partnerships are essential for innovation. Because technology is developed by universities and large companies, it is important to welcome more and more partners into the ecosystem to mine insights and better serve consumers.
4. Culture change can be achieved through empowerment (i.e., invert the hierarchy so that the front line can make decisions, which better serves consumers), collaboration, and experimentation (i.e., move quickly from ideas to viable products to scale). Leadership can establish a “learning culture” that harnesses skills and talent by recruiting, motivating, and retaining the right people.

Mr. Dale Tutt, vice president, Aerospace and Defense Industry Strategy, Siemens Digital Industries Software, explained that the cost of aircraft programs increases by a factor of 10 every 30 years. At the same time, defense programs have been on an increasing trajectory over the past 80 years in terms of how long it takes to move a program into operational capability. This unsustainable path creates a sense of urgency for transformation. He referenced the automotive industry, which has reduced its manufacturing processes (from program start to roll out) to only 2 to 3 years, as a model of achieving transformational results with digitization. He asserted that to survive disruption and thrive in the digital era, companies need to become digital enterprises—this requires taking a holistic approach that looks beyond digital tools as solutions and both changes processes and manages change.

Mr. Tutt provided an overview of Siemens’ digital journey. Its “Xcelerator” portfolio is the catalyst for the digital enterprise and is intended to scale for programs of any size. Three keys to gaining competitive advantage in this environment are the use of a comprehensive digital twin; the use of personalized, adaptable, modern solutions; and operation in a flexible, open ecosystem. A comprehensive digital twin includes the digital twin product (virtual product), digital twin production (virtual production line), and digital twin performance (real production, real product) (see Figure 3.3). He emphasized that leveraging this capability to simulate manufacturing facilities before developing them is critical to avoid potentially wasting tens of millions of dollars. Performance data can be driven back into the digital twin of the product to optimize models, and insights from performance can be used for continuous improvement on this and future products. Thus, the use of a digital twin reduces costs and schedules as well as builds momentum.

He explained that integrated solutions (e.g., model-based systems engineering, product design and engineering, verification management, integrated program planning and execution, supplier collaboration and management, intelligent manufacturing, and product support and management) accelerate programs by connecting different tools and workflows, and automating some of those workflows, to provide seamless transitions in the movement from one solution to another. These solutions are organized around five functional areas, which have varying roles throughout the product development life cycle for any program: conceptual design, product design, verification and certification, production, and deployment and sustainment. Horizontal solutions in these functional areas and vertical interconnections create a “fabric,” which makes it possible to pull verification planning, manufacturing concepts, and sustainment design forward to drive the initial design and drive program efficiencies. The concept of a digital thread for defense elevates this process by decreasing the amount of time it takes to cycle through a complete acquisition program and improving the collaboration between industry and the Air Force via better information sharing and more aligned activities. A phase in which industry is working on the program could lead to faster delivery of the capability, he continued. The objectives of this process are not only to speed up acquisition

and product delivery but also to increase operational availability and sustain effectively throughout the life cycle of a product.

Mr. Tutt shared an example of a startup company, Bye Aerospace, that has fully adopted digital—its digital enterprise provides seamless transition between design and analysis. It uses digital tools and makes connections to increase speed, thus evaluating more concepts and enhancing collaboration. With its investment in digital tools, Bye Aerospace has experienced a 60 percent reduction in engineering hours as it develops new programs. Because it is always difficult to hire the right number of people with the right expertise, this ability to move faster with fewer people is imperative, he asserted. Another company, Pilatus, has embraced digital transformation for new product development. By adopting digital tools and going faster with fewer people, it took Pilatus less than 5 years to move from program introduction to delivery of a new jet. He explained that digital transformation could also be useful for legacy programs; for example, the A-10 program transitioned from legacy, paper-based technical data and manuals to a product life cycle management-centric environment for model-based definition, configuration management, contract data management, service engineering, and wire harness design. This investment in digitization created a better understanding of the product and led to increased availability and reduced depot flow times.

Mr. Tutt offered another example of how a digital mindset can change an organization’s processes. The U.S. Navy has created digital twin models of four shipyards, and simulation is used to ensure that maintenance operations are as effective as possible now and in the future. The complex B-21 program is also making progress, using digital solutions for risk reduction. He mentioned that the first e-series aircraft, the eT-7A, embraced model-based engineering and was able to move from concept to first flight within only 3 years. A 75 percent increase in first pass quality and 80 percent fewer assembly hours were also achieved. These examples demonstrate that digital manufacturing reduces costs and increases sustainment.

In closing, Mr. Tutt underscored that digital transformation has the potential to accelerate new programs and capabilities with its focus on faster program acquisition and execution. Digital transformation can be achieved with a flexible, open ecosystem to bring technology to bear faster; improved collaboration across the value chain; and a holistic approach that considers the full life cycle early in the program as well as the roles of the people, tools, and processes.

Mr. Charles Rybeck, co-founder, Digital Mobilizations, Inc., emphasized that although Gen. Charles Brown committed the Air Force to accelerate change or lose, many misinterpreted this mandate as doing

more of what is already being done but faster and with a different name. Instead, he continued, digital transformation should be tangible, measurable, consequential, and different from “business as usual.” He asserted that the DAF has only just begun to define and respond to this challenge in ways comparable to those who have succeeded in other parts of the U.S. government, among great power competitors, and in private industry. He noted that it is difficult to sustain focused attention on any multidimensional subject; and creating an “action agenda” that prioritizes, sequences, and guides investment for an enterprise is even more challenging. However, other enterprises have proven that it is possible to overcome this barrier to digital transformation.

He described the Air Force and Space Force in the context of the National Security Enterprise (NSE), which includes DoD, the intelligence community, all associated departments and agencies, and public and private sector partners and allies. He championed joint all-domain command and control (JADC2) as the first step in becoming a true enterprise. Several years ago, a vision emerged for the NSE’s information environment to be integrated and interoperable with its operational environment; however, this did not initiate an enterprise-level digital transformation. By private industry standards, he continued, the NSE has underprioritized the need to transform enterprise-level decision making and accountabilities.

Mr. Rybeck defined digital transformation as follows: fundamentally changing the NSE’s effectiveness in prosecuting its mission to ensure strategic surprise—in other words, anticipating and overmatching potential adversaries’ capabilities and eliminating new threats. Strategic surprise is unlocked by integrating and interoperating the NSE’s information environment with its operational environment and seizing two opportunities: (1) leapfrogging competitors’ capabilities (i.e., modeling and projecting competitors’ future capabilities, taking into account their potential use of emerging but not yet operational technologies, and leapfrogging to win against these advanced capabilities instead of just overwhelming their current capabilities); and (2) mobilizing the entire enterprise’s ecosystem (including all of its human and machine assets by realigning mission partner authorities and optimizing accountabilities, drawing on each ecosystem partner’s unique capabilities to maximize the overall effectiveness and efficiency of the ecosystem). He explained that to operationalize these critical success factors, prioritized criteria are needed to predict strategic surprise effectiveness as well as maturity grids to quantify the different levels of performance capability for these criteria (i.e., decision-support benchmarks).

Despite the fact that DoD and Air Force leadership have discussed an enterprise-level digital transformation for several years, Mr. Rybeck underscored that the objectives have not been achieved. He pointed out that private sector leaders (e.g., Amazon, Apple, Google, Microsoft, Facebook) have successfully implemented platform-era technologies to merge their information environments with their operational environments. The successful enterprise visions of private sector leaders required developing both new technologies and data prioritization techniques. However, he emphasized that technologies are never a substitute for enterprise-level, proprietary strategic surprise visions. Thus, he continued, if the DAF continues to assume that supporting only one-off innovative projects will advance the mission without changing how and when investments are made in programs of record, the enterprise mission model will not evolve.

Mr. Rybeck shared three case studies that demonstrate why JADC2 has emerged as the potential “forcing function” for digital transformation, what enterprise-level decisions need to be made, and who needs to take what actions. First, he explained that the Pentagon’s Office of Net Assessment (ONA), founded in 1973, understood the need to aggregate expertise and to scope work appropriate to a challenge (i.e., “tie the ribbon properly”). However, China has been more proficient than the United States in operationalizing ONA’s insights about digital transformation.² China achieved strategic surprise on classified and unclassified fronts and announced multiple reorganizations of its own national security

___________________

² C. Rybeck, L. Cornwell, and P. Sagan, 2017, “Digital Dimension Disruption: A National Security Enterprise Response,” PRISM 7(2), https://cco.ndu.edu/PRISM-7-2/Article/1401866/a-national-security-enterprise-responsedigital-dimension-disruption/; C. Rybeck, L. Cornwell, and P. Sagan, 2018, “America’s Superpowers: How the U.S. Should Respond to China’s Informatization Strategy,” War on the Rocks, https://warontherocks.com/2018/09/applying-americas-superpowers-how-the-u-s-should-respond-to-chinas-informatization-strategy/.

enterprise. China leapfrogged the United States and its competitors with technology inspired by or stolen from the United States, and China used “informatization” as the basis for a national reorientation and a national mobilization. Therefore, he stressed that China serves as both a threat and a model to the United States.

Second, he remarked that Walmart rescoped its entire business, using the Internet to transform its relationship with its suppliers. Walmart reimagined and mobilized its ecosystems, developed interoperable collaboration software and systems to attract suppliers, and eventually became the largest retailer in the world. Walmart pioneered the use of internal cross-functional teams and ecosystem-level teams to develop ecosystem-level solutions, and it changed its business model by sharing end-consumer, point-of-sales demand data with supplier partners.

Third, the intelligence community’s Information Technology Enterprise Initiative moved the intelligence community to the cloud (using a commercial cloud services contract to make Amazon Web Services available to the entire community); however, the initiative was misunderstood and unable to gain traction. Mr. Rybeck pointed out that this example demonstrates the limitations of “tying the ribbon” too narrowly or of not having the authority to tie it at all.

He summarized that each case utilized benchmarking to design and build competitive barriers to entry, was iterative, emphasized surpassing the competition instead of striving for perfection, and began with a disruptive enterprise vision. He underscored that the most important success factor in digital transformation is “tying the ribbon properly”: the right decision scope creates the conditions for digital transformation to provide enterprise-level value. He suggested that the DAF leadership select a portfolio of transformational initiatives that are achievable within acceptable risk-return parameters.

Mr. Rybeck observed that the success factors for a broader decision scope almost always contradict the success factors used by functional siloes. Bottom-up, non-integrated, functional, siloed initiatives negatively impact the performance of other elements of an enterprise. He stated that JADC2 is necessary but not sufficient; the DAF needs mechanisms to operationalize its objective. In closing, he advocated for the DAF to use fifth generation decision support to align, orchestrate, and instrument digital transformation solutions by tying the ribbon properly, doing what is necessary to use those decision support benchmarks, and chartering a rigorous and expedited net assessment of enterprise-level digital transformation. This net assessment, he continued, should include comprehensive benchmarking to illuminate a path for the enterprise to prioritize, sequence, and invest, with the purpose of leveraging the Advanced Battle Management System and JADC2 to ensure strategic surprise success against competitors and adversaries.

Open Discussion

A participant inquired about cross-sector partnerships and how to establish a baseline level of understanding. Mr. Welde explained that establishing a baseline and the returns on investment is a collaborative activity; the partners co-create a brief that describes the task at hand. Alignment is feasible when partners are involved upfront in the process of innovation.

Another participant asked which users the Air Force should be targeting in its digital transformation. Mr. Rybeck noted that focusing only on the end user (e.g., the fighter pilot) is too limited; it is also critical to focus on the adversary’s use of new capabilities in the physical and cyber realms. He added that it is important to consider man-machine collaboration instead of thinking only about individual human beings as end users. Mr. Rybeck and Dr. Drew suggested that it is not a matter of identifying all of the actual end users but rather locating the actionable decision makers at a proper scope that can increase efficiency related to digital transformation.

In response to a participant’s question about how to balance an enterprise approach with a “quick wins” approach, Mr. Rybeck said that the two approaches could be combined. He emphasized that although quick wins are instructive and help illuminate the right path, incremental progress is not transformational; if too much credit is given to people in an organization for making small changes to existing processes, there is no motivation to make large-scale changes. He noted that the Air Force and Space Force currently have a

low tolerance for failure, and the initiatives presented all claim success. He advocated for more honest conversation, frank assessment, and lessons learned as a “quick win.”

Dr. Annie Green, data governance specialist, George Mason University, asked Mr. Rybeck how levels of abstraction (e.g., processes, products, and analytics) are delineated and integrated to accommodate scope, goals, and objectives in JADC2. He noted that JADC2 has the advantage of being focused on next-generation capabilities. Next, leaders need to consider how the Joint Requirements Oversight Council could operationalize the construct. The simple approach, he continued, is to compare the packages of capabilities to those of the adversaries in a classified environment to confirm whether the United States is leapfrogging their capabilities.

Lt. Gen. Hamel asked the panelists about their experiences of achieving digital transformation in the business sector with both bottom-up (i.e., innovation, digital infrastructure, data, people, process) and top-down (i.e., vision and strategy) approaches. Mr. Welde said that although digital transformation starts with the top-down vision and strategy, the bottom-up approach plays an equally important role, as digital transformation cannot be executed in a vacuum. He stressed that a grand vision for transformation is useless without clearly articulated milestones and measurable objectives, and he described the biggest challenge as achieving end-to-end integration across an organization. Mr. Tutt echoed Mr. Welde’s observations and reiterated that the top-down strategy is needed to guide the organization but the bottom-up perspective provides the buy-in that is critical to success. He emphasized that all members of the organization need to feel comfortable that their jobs are secure, that their jobs could improve or become easier, and that they will be empowered to solve problems creatively.

Dr. Drew supported the notion that change management is critical to the success of any digital transformation, and she invited the panelists to share related lessons learned. Mr. Rybeck noted that once people understand how different “business as usual” is from digital transformation, it is possible to apply standards to measure whether the right team is in place and the right conversations are occurring. In a transformative environment, he continued, people should be evaluated based on their contributions to the enterprise vision, and it is important for champions to provide top-cover for people doing innovative work. Mr. Welde observed that successful leadership and execution of a digital transformation requires a specific type of expertise. In any transformational effort, people want to see results quickly, but it is important for change leaders to calibrate decisions carefully: not too fast so as to leave the organization behind, but not too slow so that the competition moves ahead. Leadership is also responsible for ensuring that individual initiatives are moving in unison toward the “big picture.” He emphasized that change management is foundational to transformation. Mr. Tutt explained that the success of a change is the product of the acceptance and the quality of the change. A change champion expresses and communicates a vision continuously, with an achievable objective and quick wins.

Mr. Alden Munson, senior fellow and member, Board of Regents, Potomac Institute for Policy Studies, asked whether industry experiences could realistically be applied to government institutions. Dr. Drew acknowledged that the basic tenets of process improvement and efficiency apply to both domains. Mr. Tutt added that there are several similarities between the domains. For example, in both industry and government, people are continually changing roles, and arriving and departing, over the course of a long-term initiative; managing that change appropriately is critical. The notion of “profitability” in industry is also applicable to DoD, which deals with taxpayers and ever-shrinking defense budgets that affect national security.

Dr. Rama Chellappa, Bloomberg Distinguished Professor of electrical and computer engineering and biomedical engineering, Johns Hopkins University, wondered if people have misinterpreted digitization as government surveillance. Mr. Welde suggested that this concern about digitization could be reduced if leaders addressed safety and security, thus removing any perceived threats, and communicated the principled uses for data analytics.

LESSONS FROM OTHERS “LIKE THE AIR FORCE”: PANEL TWO

Dr. Jay Walsh, interim vice president for economic development and innovation in the University of Illinois system, described his time as vice president for research at Northwestern University, where he supported efforts analogous to those found in industry: he oversaw 2,000 “small businesses” (run by faculty), each of which wrote its own “business plan” (i.e., proposals), ran its own operations, cultivated a strategic vision, created “revenue” (i.e., grants), and developed a “product” (i.e., new knowledge and graduates). To facilitate this amount of research, the university needed to be agile, provide a service for its “customers” (i.e., the researchers) that was available 24/7, and enable self-service. Because each researcher ran a slightly different “business,” each needed slightly different services. During Dr. Walsh’s 12-year tenure, the university enabled more research by significantly increasing its IT infrastructure with the purchase of one new system each year.

He described a major initiative in Northwestern University’s healthcare system to create an electronic data warehouse. About 20 years ago, the notion of connecting genomic data to healthcare data was nascent, and healthcare data were not readily accessible to researchers because the hospital owned those data. It was particularly challenging to develop a system that allowed the university’s researchers to access clinical data that had been generated in collaboration with physicians. Approximately 14 years ago, an agreement was reached that allowed the university to move the data from the electronic healthcare system into a research system. That system now contains data from 17 million patients, and each night the system moves ~3 billion data elements from the healthcare systems into the electronic data warehouse from ~140 sources. He explained that significant synchronization enables this process each night. The electronic data warehouse now serves as a comprehensive repository containing all of the clinical and research data associated with the university’s medical school—an effort that was driven by the researchers. Although the hospital was initially not very interested in this activity, its perspective has changed because the incoming data include operations and financial data. As a result, this database is valuable for patient care, healthcare operations, and medical education: the hospital can provide better care for patients, and the medical school can advance research and train the next generation of clinicians. He designated this case study as applicable to the DAF because if the Air Force, Space Force, and other organizations combine their data, they could analyze the relationship among those data and determine how to operate better and achieve the mission.

Dr. Walsh emphasized the value of researchers having access to data but noted that the regulatory environment presents challenges. Data security was built into the electronic data warehouse from the beginning to protect identities, although this was difficult because the 2,000 “companies” (faculty members) worked independently. Governance and policies were needed to address these challenges, which are similar to those faced by the Air Force and Space Force. Building this system at the university also included creating a system of support people to help researchers access and operate the system so as to be able to collect and analyze data—an action that contributed to the success of this transformation.

Another major initiative related to moving large quantities of data. Whereas healthcare data are largely confined geographically in hospitals and do not have to be moved long distances, Dr. Walsh remarked that other types of data, such as astronomy data, are much more difficult to move. Yet large quantities of data have to be moved around the planet in order for astronomers to compute and analyze those data. The AI stack has at its foundation computing and devices that collect data; once the data are on a computer, it is possible to do machine learning (ML) and decision support to eventually reach a level of autonomy. Academic researchers have built a network of fibers across the globe to enable this work, which will accelerate the innovation and advanced global communications of the Internet. He clarified that these are mostly private fibers used for the explicit purpose of academic research (the use of commercial fiber was too expensive). Academia drove the development of these fiber-based systems and related components because they could not do their jobs without them. Companies are primarily directing advances in computing, although research laboratories and universities are also driving new ways of computing, new devices, and new mechanisms to gather and move data.

In closing, Dr. Walsh explained that a university, with its varied faculties and laboratories, is a decentralized system that needs centralized services to operate. This situation parallels that of industry and

uses a similarly consumer-centric, program-driven approach. If any part of the ecosystem (e.g., computers, algorithms, models, devices, human support) were missing, the ecosystem would not be functional.

Ms. Margaret Palmieri, Special Assistant to the Vice Chief of Naval Operations, U.S. Navy, said that the Navy’s journey of digital transformation began in 2016, although the Navy warfare community has been on a mission to connect weapons sensors and shooters since 2000. She considered cultural issues to be the primary barrier to realizing this goal, as well as organizational challenges such as budgeting processes.

She explained that the Navy has four distinct communities: surface warfare, submarine warfare, aviation warfare, and information warfare. The Navy recognized the need to transition into a more multi-domain, cross-domain service, which could be enabled by digital transformation. A taskforce including representatives from across the Navy—systems commands, program offices, engineers, requirements officers, and members of the research and development community—was stood up in 2016. Recognizing this as a cross-functional problem, as opposed to one only for the requirements or acquisition community, a Digital Warfare Office (DWO) materialized, which included detailees from the engineering community. DWO aligned its efforts with the operational concept Distributed Maritime Operations. Other organizations were benchmarked and consulting literature on successful digital transformation was reviewed, revealing opportunities both to improve warfighting systems with better networking and integration and to improve readiness with better use of data. DWO thus stood up two lines of effort: (1) using data and analytics to better support decision making to increase the readiness of aircraft, ships, and submarines; and (2) creating an architecture and a data approach for warfare to connect weapons sensors and shooters. DWO used a “bimodal approach” to garner quick wins, generate buy-in, and learn through pilot projects—five readiness pilots and five warfighting pilots were created, and a blueprint for a future architecture was developed. This architecture was envisioned to be more agile than the vertically integrated, tightly coupled, single-purpose systems in existence and to allow the implementation of needed software approaches (e.g., containerization). After a few years, when decisions needed to be made about the architecture, Ms. Palmieri credited her team for their assistance, noting that the best leadership has expertise in change management and is technically savvy. She also described the Navy’s creation of Project Overmatch in 2019 to address the warfare effort and its development of the Perform-to-Plan process (“Get Real, Get Better”), where data and analytics are used to identify readiness challenges.

Ms. Palmieri detailed some of the successes of the Navy’s early efforts in digital transformation. She and her team were able to identify the barriers to connecting weapons sensors and shooters: culture, processes, and organization were the problem, not technology. There was also a realization that a senior-level leader who understood and was willing to remove those barriers was needed. Another success was that the transformation with structured with a mission focus: all initiatives had an associated mission outcome. In addition to developing the architecture and trying to integrate systems, the team worked toward having these systems speak the same language, which made it possible to create standard definitions for the requirements process. She emphasized that even though transformation work is highly mission focused, it also needs to be highly collaborative.

She noted that the transformation effort was initially too focused on systems engineering (i.e., trying to pursue the “perfect” architecture) and not focused enough on software. An important lesson learned was to balance the architecture for current and future systems with the agility of add-ons for immediate implementation. She added that the team was able to gain more momentum, moving from pilot to scale, when it could reduce risk by providing the needed talent to other mission leaders for their projects. A new unmanned system cross-functional team works to minimize the handoffs among requirements, acquisition, experimentation, and the fleet, aligning everyone on the same team, with a 4-star leader to help remove barriers.

In closing, Ms. Palmieri shared key takeaways for successful digital transformation: define the outcome and work backward, find cross-functional collaborative leaders with technical fluency, have consistency in goals to overcome turnover in military leadership, balance centralized approaches with more decentralized approaches, expect to be surprised during the journey, harness the talent of digital natives, make learning

and risk-taking acceptable, and determine who is going to eliminate barriers and translate mission to technology. Dr. Drew emphasized how many of these lessons learned apply directly to the Air Force.

Dr. Jan Neumann, executive director of machine learning, Comcast, said that despite the differences between industry and the military, several principles and lessons learned apply. He explained that the Applied AI and Discovery Group that he leads at Comcast has a mission to “delight [its] customers by connecting them to the moments that matter.” The group’s approach is to eliminate intermediaries and connect directly with the customers, and it is responsible for voice control from the remote, content discovery (i.e., recommending the right show at the right time), media analytics, digital home (e.g., smart cameras, door sensors, temperature sensors), and customer experience and customer service (i.e., proactively solving billing and repair issues). Underlying all of these services are platforms for experimentation, data, and AI and ML. He emphasized that customer feedback is essential to understanding product efficacy—digital transformation makes it possible to connect directly with customers to get feedback more regularly and use automation to accelerate speed to value. He noted that it is challenging to translate complex problems into programs rapidly. However, ML can be used to create a nested feedback loop to learn faster and provide solutions more quickly. For example, natural language processing can be used to translate natural speech into words with higher accuracy than humans.

Dr. Neumann shared a use case about the digital transformation of Comcast’s customer service. Comcast built AI- and ML-driven assistance for its customer service agents so that data from all agents across geographic locations could be aggregated for collective learning and for quicker problem solving. In some cases, with the use of predictive algorithms, customers will no longer need to interact with an agent at all (e.g., if a customer enters Comcast’s app to troubleshoot a slowdown in an Internet connection, the interface will be adjusted based on that specific context). This makes it possible both for the customer to address the problem more quickly, and for Comcast to learn whether its predictions are accurate and if changes are needed. Comcast is also working to be more proactive: it has the ability to monitor for issues and send a repair truck before a customer is aware of a problem. These services are enabled by the Xfinity Assistant, which can be accessed via Facebook Messenger, general apps, the television, the website, or an Apple watch. If customers type a problem into the Assistant, ML algorithms are used to understand what type of problem exists and offer the best troubleshooting solution, only escalating to a human agent when needed. This reduces time and streamlines the customer experience. It also offers a way to measure outcomes and improve, with continual iteration of the models. He explained that the feedback loops are used to optimize decision making, detect anomalies, and customize entertainment based on what customers need and what is known about them: faster feedback cycles create better, more personalized customer experiences (see Figure 3.4).

Dr. Neumann described several lessons learned during Comcast’s journey to improve customer experiences. It is important to focus on the fundamentals in order to make sense of the data by using the digital hierarchy of needs. First, information is logged and collected from applications. The next step is to build an infrastructure that can move and store the data. It is at this stage that issues of security, privacy, and access become critical, as the data are now ready to be explored and transformed. The next step is to provide accurate labels to ensure that the systems can learn, and then use simple ML or optimization algorithms (or, if necessary and if enabled by the data, AI and reinforcement learning) to find solutions. Another important lesson learned is the importance of creating loosely coupled yet highly aligned teams focused on products (not projects) that improve customer experience. It is critical to establish the right incentives as well as to develop appropriate metrics as indicators for the desired outcomes. He suggested that the best aligned incentives emerge in teams that are responsible for both costs and benefits. He emphasized the value of experimentation and using data for decision making, as well as building platforms that make it easier to test hypotheses and to iterate faster. Software and data science are creative acts with high variability, which makes estimation difficult. He advocated for less time spent planning and more time working through a priority list.

He noted that Comcast developed a Center of Excellence to establish best practices. Self-service data and compute platforms codified these best practices to make it easier for everyone to access the data and compute needed to solve their problems. Data scientists and ML researchers from the Center of Excellence

were embedded with the product and business teams to better combine technical and domain knowledge. Change management is crucial for transformation, he continued, because new approaches can be viewed as threatening to teams. He suggested that organizations implement the ADKAR (Awareness, Desire, Knowledge, Ability, Reinforcement) model for change management. Retaining the human-in-the-loop also helps with the adoption of a transformation, with a focus on “assistance” over “automation,” because the human experience is key to solving problems where data do not tell the whole story. Last, he advocated for beginning the process using the rules of existing code as baselines, setting up data pipelines and storage, and then improving upon them using ML.

Open Discussion

Dr. Drew observed several common themes throughout the presentations in this session: experimentation, change management, faster learning, and AI. She wondered which lessons learned from the development of DWO could be most valuable for the Air Force. Lt. Gen. Hamel noted that each service has a unique, long-standing organizational structure and culture. For example, the Navy has well-established warfare communities among surface, subsurface, aviation, and information operations, and the Air Force has more platform- and mission-oriented alignments with Air Combat Command, Mobility Command, and Special Operations. The Air Force tends to be organized around functions more than communities, and he hypothesized that this influences how well initiatives work in one service versus another. Gen. Holmes mentioned that the functions in the Air Force model are united in that all expect centralized control and decentralized execution through an operations center. Because a key aspect of digital modernization is recreating that command-and-control process, he suggested that the Air Force focus on unification at the command-and-control level instead of within the communities. Lt. Gen. Hamel posited that the Air Force has an advantage over some of the other services in that it has an experimentation center that crosscuts every mission area. Gen. Holmes added that while the Navy has a long tradition of tailoring individual ships, the Air Force often applies more standardization across a platform. The Navy seems to be

more willing to experiment, he continued, perhaps because a ship captain has more independence than his Air Force counterpart.

Lt. Gen. Hamel asked what resources (i.e., talent and dollars) are needed to achieve the transformations described by the panelists. Ms. Palmieri replied that it depends on whether an organization’s execution model is decentralized or centralized. For example, even though the Navy created DWO, the digital transformation was mandated by the Chief of Naval Operations (which is equivalent to commander’s intent in the Air Force), so many individual commands stood up their own efforts with decentralized resourcing. DWO pursued a new funding line to obtain the right expertise in data analytics and network design to integrate the communities. Individual commands were also encouraged to contribute to DWO’s pilots. However, gaining the support of the authorizers does not necessarily translate to increased budgets; she proposed clearly articulating what will be produced and how the funding is directly linked to achieving a specific outcome. Dr. Walsh described securing talent and money as a challenge for every organization. In academia, it is important to identify the mission, determine how a digital transformation will help achieve the mission, and communicate clearly to the administration and the board of trustees how their resources will enable the mission. Dr. Neumann added that the difficulty of securing funds correlates to the amount of money needed. Detailed business development is expected for large investments. He explained that it is crucial to calculate and identify potential payoff versus risk when seeking investment; in industry, if a company does not change, it will be left behind its competitors (e.g., Netflix, Roku).

Mr. Munson wondered about the security of Northwestern’s data warehouse. Dr. Walsh responded that it is connected to the Internet, but a more secure system with fewer connections would not be as useful for researchers. The data warehouse is located behind a robust firewall, and the university stood up an organization to ensure that everyone follows the rules for access and device registration. Mr. Munson questioned the applicability of this use case to the Air Force; he expressed concern that an advanced persistent threat could breach that level of security. Dr. Walsh reiterated that the data are well protected and that there is a team working on defenses; however, he emphasized that anything electronic could always be vulnerable. Dr. Drew pointed out that when disparate data sources are brought together in any enterprise and used in unanticipated ways, unexpected capabilities emerge. Thus, classified entities may be created, which is an important security issue for the Air Force to consider.