2
Workshop One, Part Two
OPENING REMARKS
Workshop Series chair Ms. Deborah Westphal, chairman of the board, Toffler Associates, highlighted the value that the previous workshop speakers placed on strong leadership in the U.S. Air Force (USAF). In this time of rapid technology development, she wondered whether the USAF is paying enough attention to what the warfighter of the future would look like—in terms of technology, platform, and human transformations. Understanding the roles of people and concepts of operations is a central purpose of this workshop series, she continued. She questioned whether the USAF is investing enough in its people, leadership, and thought processes to succeed in the future. Gen. Gregory “Speedy” Martin (USAF, ret.), GS Martin Consulting, Inc., agreed that people are key to the USAF’s success. Reflecting on Ms. Natalie Crawford’s presentation from the previous day of the workshop (see Chapter 1), he reiterated the value of building relationships with and trust in people who are personally committed to the organization and exercise critical thinking skills to help solve problems. He cautioned about the dangers of the USAF’s insularity and mentioned that the U.S. Navy has been very successful in its outreach to industry and strategic thinkers for fresh ideas. Sometimes airmen become self-dependent and too narrow in their focus, he continued—it is important to convince USAF leaders of the value of spending time outside of the organization talking to trusted experts who can help make a difference. Recalling the workshop planning committee’s discussions prior to the workshop (see Preface), Ms. Westphal also wondered whether the USAF has become too business-focused (e.g., on technology, platforms, and acquisition) and is losing the warfighting mentality.
Dr. Julie Ryan, chief executive officer, Wyndrose Technical Group, observed that all of the presentations from the first day of the workshop series stressed the importance of data. Effects are realized when data are exchanged; however, too many blockages to data transmission exist, including in the realm of human-to-human data transfer. Dr. Joseph “Jae” Engelbrecht, president and chief executive officer, Engelbrecht Associates, LLC, shared his summary of the first day of the workshop series with participants (see Box 2.1).
TIME CYCLE REDUCTION IN THE DEVELOPING AND TRANSITIONING OFNEW TECHNOLOGICAL CAPABILITIES: A SENIOR LEADERSHIP PERSPECTIVE
Gen. Lester L. Lyles (USAF, ret.), former Vice Chief of Staff, USAF; Commander, Air Force Materiel Command; director of the Ballistic Missile Defense Organization; director of the Tactical Aircraft Systems Air Force Sustainment Center; and Short-Range Attack Missile Program Element Monitor, presented several acquisition
success stories that could apply to non-acquisition time-cycle reductions for day-to-day efforts to get technologies in the hands of the warfighters. He advised careful thought when there is a need to accelerate and advocated for open communications between program directors and Congress.
He recalled that from 1978 to 1981, the position of Assistant Secretary of the USAF for Acquisition, Technology, and Logistics did not exist: Acquisitions, science, and technology were all under the purview of Gen. Alton Slay, Commander of the Air Force Systems Command. Gen. Slay embraced development planning as a means to understand technologies, work with the warfighter, conduct early-stage experimentation and prototyping, identify successes, and aggressively try to get technologies into the hands of the warfighter.
Gen. Lyles turned to a discussion of his involvement in developing the F-16 C and D aircraft and transferring those capabilities to the warfighter. The Multi-Stage Improvement Program (MSIP), the umbrella program for the F-16 C and D aircraft, focused on creating incremental improvements in capabilities and fielding them to warfighters as quickly as possible. Block upgrades on the F-16 were an essential part of development planning, he explained, as new capabilities were implemented on 2- or 3-year cycles. Congress, the USAF leadership, and the Major Commands (MAJCOMs) bought-in to MSIP, and budgets were aligned so that programs could mature in time to go into the F-16 when the block improvement was needed. He noted that the Block 60 F-16, which includes the electronically steered array radar system, is now owned by the United Arab Emirates, a U.S. ally. Gen. Lyles described the F-16 program as a great success; it demonstrated the value of block upgrades and of pacing improvements according to both the budget and the warfighter’s needs. The F-16 program also underscored the value of strategic planning: In 1984, the Falcon Sentry program devised a 16-year plan of upgrades and technology needs for the warfighter of 2000. Gen. Lyles suggested implementing that level of strategic planning for weapons systems; it is not yet occurring because budgets are constrained and program directors are not given the freedom to think ahead.
Offering a tactical perspective, Gen. Lyles provided another example of getting capabilities into the hands of the warfighter quickly. The High Gear programs1 addressed the warfighter’s most immediate needs—within 12–18 months. In order to complete the incremental improvements that would address these short-term warfighter needs, Gen. Ronald Yates established agreements with each MAJCOM without going through a program of record, which would have taken several years. Gen. Lyles recruited experienced personnel from the laboratory and product centers (both civilians and military) to help make these efforts successful. He observed that High Gear, similar to AFWERX, focused on changing the culture of innovation in relation to acquisition.
Gen. Lyles explained that the 2014 National Research Council report Development Planning: A Strategic Approach to Future Air Force Capabilities2 was an effort to revitalize development planning in the USAF. He emphasized the significance of support from USAF senior leadership; the Chief and Secretary enabled the development planning effort. He also referenced the 2016 National Academies of Sciences, Engineering, and Medicine report The Role of Experimentation Campaigns in the Air Force Innovation Life Cycle,3 which discussed development planning in relation to the Vanguard program. The report looked specifically at the organizational structure needed to support experimentation (e.g., the Air Force Research Laboratory [AFRL], Air Force Materiel Command, and the Air Force Warfighting Integration Capability [AFWIC], now called AF Futures). Vanguard development planning was effective in part owing to the efforts of senior leadership—the MAJCOMs supported, embraced, and depended on Air Force Systems Command (AFSC) and the Vanguard program4 to help understand what to focus on for the path forward. It is important to maintain a connection between the technology community and the warfighter, and Gen. Lyles expressed concern about how many responsibilities now fall under AFWIC. The technologists in AFRL no longer have the
___________________
1 For a brief description of High Gear, see National Academies of Sciences, Engineering, and Medicine, 2016, The Role of Experimentation Campaigns in the Air Force Innovation Life Cycle, The National Academies Press, Washington, DC, Recommendation 6, p. 68, https://doi.org/10.17226/23676.
2 National Research Council (NRC), 2014, Development Planning: A Strategic Approach to Future Air Force Capabilities, The National Academies Press, Washington, DC, https://doi.org/10.17226/18971.
3 National Academies of Sciences, Engineering, and Medicine, 2016, The Role of Experimentation Campaigns in the Air Force Innovation Life Cycle.
4 The Vanguard program cited here, discussed extensively in the 2014 Developmental Planning NRC report, is not the same as the current AFRL Vanguard program.
strong connection to the warfighter that they had in the earlier days of development planning, he continued. Since AFWIC is responsible for integration, Gen. Lyles asserted that it could work hand-in-glove with AFRL.
In closing, Gen. Lyles provided examples of efforts to modernize and change culture from the United Services Automobile Association (USAA). USAA is an organization that works continuously to reinforce its mission and vision. It has dedicated $100 million/year for innovation and allocates additional funds to adapt technologies to address issues of compliance and increase automation. USAA recognized that it was necessary to transform technologies to better serve its members and to remain relevant, and it established innovation laboratories focused on insurance, banking, and real estate for the future. It also created a program called “Everybody is an Innovator,” which attempts to change the culture of the 33,000-person workforce. This program encourages any and all employees to share their innovative ideas with the company.
In response to a question from Gen. Martin about changing the culture within the USAF, Gen. Lyles stated that the most significant challenge was helping people to realize that they were all part of one team within the Air Force Materiel Command working to support the warfighters. It is also critical that people are comfortable with experimentation and unafraid of failure. Ms. Westphal asked if the USAF is doing as well as USAA in inviting conversations about innovation. Gen. Lyles explained that the goal of AFWERX is to foster a culture of innovation within the service. He described it as the right approach, but whether it will be successful remains to be seen. Ms. Westphal pointed out that although a desire for cultural change may exist, an adversarial attitude toward “outsiders” with novel ideas still exists within the USAF. Gen. Lyles noted that many of the plans—for example, engaging small business, universities, and minority-serving institutions, and holding pitch days—are moving in the right direction, but those strategies have to be executed and institutionalized.
Lt. Gen. Ted Bowlds (USAF, ret.), chief technology officer, IAI North America, asked how to ensure that an innovative idea moves from the laboratory to the warfighter instead of to the valley of death. Gen. Lyles commented that AFWIC is expected to serve as the voice of the warfighter and to help define requirements for technologies from the laboratory or in the Air Force Materiel Command; if AFWIC succeeds in this role, it is possible to bridge the valley of death. Gen. Lyles added that Vanguard (the current AFRL program) sets aside 20 percent of 6.2 funds for high-payoff programs that could address game-changing capabilities. A purpose of the Vanguard program is to maintain a close connection to the warfighter, to secure early buy-in, and to avoid the valley of death. Three Vanguards that address different domains are active now; however, an objective of the Vanguard program is to establish programs of record, which creates a bureaucracy that adds another valley of death and could delay getting the capability into the hands of the warfighter. Gen. Lyles suggested the development of a different approach for the Vanguard program that would allow for rapid transition of capabilities to the warfighter. Dr. Engelbrecht inquired about how commercial organizations are transforming and innovating. Gen. Lyles noted that even though the language may be different, there is a parallel between what is being discussed for the USAF and what is happening in the commercial world.
CUTTING THE GORDIAN KNOT: EXECUTING TIMELY COMBAT-FOCUSED RESEARCH, DEVELOPMENT, TEST, AND EVALUATION AND RAPID DEPLOYMENT VIA SMALL CREATIVE TEAMS AND ORGANIZATIONS
Dr. James Roche, former Secretary, USAF, explained that his experience in the U.S. Navy and in industry motivated his interest in increasing the speed of USAF processes. He emphasized that the USAF’s practice of creating programs of record before funds can be allocated and progress made leads to failure in a dynamic environment. He expressed his frustration about the difficulty of making progress when he arrived at the USAF. For example, he had to fight to build all of the Predators with capabilities for both weapon use and reconnaissance instead of building two different Predators. He eventually accomplished this objective, developing strong relationships throughout the process. Dr. Roche also described his experience with the B-1, which, because it was never modernized, was known for its inaccuracy. He was petitioned to eliminate all of the B-1s, but instead he eliminated only half (turning those into spare parts lockers) and worked on improving the remaining aircraft. He explained that the USAF had not been studying the inherent capabilities and comparative advantages of its systems—only a few changes to the aircraft were needed to make substantial improvements. Sniper pods were added, and when Sniper became a program of record, the Lightning program was in danger of being eliminated. He proposed maintaining Sniper
and Lightning, which actually decreased costs over time and allowed for improvements within both programs. He emphasized that the first step to addressing a problem with an aircraft is to simply look at the plane differently than in the past—a key tenet of successful integration.
Dr. Roche described his role as a sponsor of the Rapid Capabilities Office (RCO), an organization of 200 people, approximately 160 of which are talented active duty USAF officers. The RCO is a model of how initiatives led by dedicated teams with fresh ideas can lead to tremendous progress. He explained that the RCO was initially structured as a flat organization focused on innovation and speed. RCO officers have undertaken a massive challenge in running the B-21 program; however, because some parts of the B-21 are linked to the bureaucracy, it is taking longer than expected. He advocated instead for getting the capability in the hands of the warfighter first and addressing problems as they arise later, an approach that could be aided by artificial intelligence (AI). He portrayed the RCO as “wonderfully disrespectful,” in that young officers are not afraid to challenge their superiors and offer new ideas. It is acceptable to fail in the RCO, as long as something can be learned from the failure. He expressed disappointment with the USAF’s limited efforts to utilize and promote such talented individuals just because they do not have experience as pilots.
Before concluding his presentation, Dr. Roche shared an anecdote about the realization that a capability (the ROVER) was needed in the cockpit to show a pilot what the Predator sees, which would allow the operator of the gunship and the Predator to be able to communicate. What is important is that this improvement did not come up through a system—in other words, it did not have a review at the Pentagon. He also discussed small improvements that were made to the X-37B, leading to its broader acceptance. He emphasized the value of using real data on equipment operating in space to understand whether or not it is working. He also suggested that the use of individual software be prohibited; open architecture is necessary for people to be able to “plug-in.”
During the question-and-answer session, Gen. Martin described that when a capable organization gets locked into working with certain contractors and draws congressional support, the environment of innovation disintegrates because the organization becomes self-protective. He wondered how to prevent that scenario. Dr. Roche replied that the RCO gets to make decisions about the companies with which it works. However, he said that although the debate over the systems can be controlled, the bureaucracy that inherits the systems cannot be controlled, which leads to delays. He expressed frustration with the “burdensome system” of the Pentagon, which often halts entire programs when only a simple repair is needed. He asserted that the acquisition system has to change and noted that the RCO has the ability to exchange perspectives between operators and scientists. He championed borrowing from others to bring real capabilities to different parts of the USAF, often at a low price. Dr. Roche was inspired by the level of talent within the RCO; for example, the officers discovered a safe way to land in Afghanistan simply by changing the process of departure and arrival. When there is a receptive audience, much can be achieved. The most important part of warfare, he continued, is the mental component.
Ms. Westphal questioned whether there was anything that Dr. Roche was unable to change during his tenure with the USAF. Dr. Roche (and those before him) had a goal to fix the USAF Academy and to address an ongoing problem with tankers; however, the tanker problem still exists 18 years later. Because the Dutch have a tanker with a boom and a camera that has worked well for 20 years, he wondered why the United States does not buy theirs. He added that he has two rules for success: be honest and always keep your word. Hon. F. Whitten Peters, senior counsel, Williams and Connolly, LLP, asked about the recruitment process for RCO, and Dr. Roche replied that staff are recruited from the science and engineering community. There is still an antiquated rule that prevents the young men and women from going to graduate school without a certain number of years of service, which deters many from participating. If the USAF eliminated certain “rules” and made better career opportunities available, he continued, more talented young men and women might stay in the service.
THE EMERGENT DIGITAL THREAD AND DIGITAL TWIN: IMPLICATIONS FOR COMBAT DIAGNOSTICS AND PROGNOSTICS
Dr. Thomas Hedberg, Jr., mission lead, Acquisition and Industrial Security, Applied Research Laboratory for Intelligence and Security (ARLIS);5 visiting associate research engineer, Institute for Systems Research; and
___________________
5 For more information about ARLIS, see University of Maryland, 2020, ARLIS, http://arlis.umd.edu, accessed September 17, 2020.
associate research engineer, University of Maryland, College Park, remarked that the USAF is “best in class” in digital innovation but has failed to communicate its success stories or a cohesive strategy. (The U.S. Navy tends to be more successful in communicating its strategy.) Adversaries are using blended methods to disrupt and compromise U.S. assets, capabilities, and supply chains while advancing their own interests. He explained that a cohesive strategy would include how to be more agile more quickly and how to better link across supply chains, as well as consideration for security costs. He noted that the USAF would benefit from a clear, integrated strategy for deploying systems and capabilities with critical playbooks (e.g., in zero trust architectures for microelectronics) that enable rapid response and quick wins in specific areas.
Dr. Hedberg described a model-based enterprise (MBE) as an organization, or set of organizations, that uses digital representations as the communication medium and authority in an industrial endeavor. Example classes of models include the system model, product model, and process model—the digital thread and the digital twin connect those models together in an MBE. He defined digital thread as an integrated information flow that connects all the phases of the product life cycle using an accepted authoritative data source. He defined digital twin as an integrated model, enabled by the digital thread, that combines data from both the cyber space and the physical space to mirror and predict things over the life of the model’s corresponding physical twin.
Dr. Hedberg portrayed supply chains as sociotechnical systems made up of people, technologies, organization, and management, including policy (see Figure 2.1). The internal system of those personnel, technological, and organizational management subsystems makes up the supply chains, which are under constant attack. The Department of Defense (DoD) is trying to defend and mitigate against such attacks and to make observations of the system to apply control to it—as a result, research, development, testing, and evaluation are emerging.
In order to reduce the time cycle to field systems and capabilities, one has to understand what is interacting inside the system, how the personnel are interacting with the technology, and how the overall system interacts with its external environment, Dr. Hedberg explained. He said that a system’s behavior is a function of that system’s interactions with its environment. The theories of joint causation, optimization, and design have to be respected—when one part of the system is changed, changes are caused in other parts of the system; when one part of the system is optimized, optimization in other parts have to be considered; and these connections have to
be considered in terms of how everything works together. Holistic systems thinking and systematic analysis is crucial, he continued, making way for MBE thought processes.
Dr. Hedberg provided an overview of several National Institute of Standards and Technology (NIST) MBE Summit presentations from the past 10 years. The USAF’s 2011 presentation focused on identifying the costs—billions of dollars of lost opportunity or cost to programs—associated with a lack of interoperability. This highlighted the need for the USAF to move to a data-driven environment and establish a cradle-to-cradle digital thread, which could link systems to share information effectively and efficiently. Focusing on the need to provide information to decision makers, in 2013 the USAF pursued a human-based digital enterprise that would be able to take data from all domains and all phases of the life cycle and present them to a human in a way that could be understood, interpreted, and used for decision making. This is when the three-dimensional (3D) technical data package emerged; however, little additional progress has been made. In 2016, the USAF concentrated on the need for digital artifacts to achieve model-centric engineering and to utilize a digital surrogate for decision making, although several barriers remain. Deciding whether to go through incremental or adaptive innovation or to strive for the stepwise function to achieve some discontinuous jumps in innovation is key to understanding how to move quickly in this area. In response to this concern, the USAF formed a Digital Enterprise Office in 2019. Dr. Hedberg said that the next step is to think about integration—how to move to a digital engineering project that is linked to the logistics and sustainment project to ensure that there are proper inputs and outputs and efficient flow of information. Initial progress has been made with the development of a roadmap and action plans.
For 2020 and beyond, the USAF is exploring the notion of “digital twin meets autonomy.” He stressed that the USAF has to focus on integration (i.e., integrating systems engineering with design, sustainment, the warfighter, and the policy makers) and holistic decision making (i.e., decisions can be broken down into actionable tasks that the individual units can execute). All of the work in autonomy could immediately leverage digital twin activities, he continued, but there is little discussion about how this could be done at mass scale. He explained that to move forward in the future environment, the USAF has to better manage its operations of system engineering, trade studies, design, manufacturing, and sustainment, perhaps via the implementation of a common workflow model. He reiterated the importance of integrated logistics. Proper planning and prognostics of understanding different scenarios and what-if analyses of a particular engagement matter. Whether the conflict is physical or cyber, the key is to understand it systematically—logistics, transportation, capabilities, and economics. Dr. Hedberg observed that combat readiness is both a national defense problem and an economic security challenge: If the United States does not have the manufacturers who make up the defense industrial base available and capable to produce systems, it does not have the appropriate level of readiness.
Dr. Hedberg noted that the Connection Interoperability Paradox emerges in a digital enterprise. A top-down, bottom-up, middle-out approach would work best. A problem that surfaces is a weakness in trusted decision making in distributed environments. Standards such as ISA-95 can segment different levels, creating artificial boundaries and barriers. While ISA-95 helps to identify where new technologies are needed, he asserted that it is not useful for making enterprise decisions. There is no clear mapping of how data flow up and down efficiently and effectively in such a framework. This also leads to the creation of more artificial cylinders of excellence. Horizontal and vertical integration across the enterprise does not work in the current environment because each level has a different language and approach, which creates both a military problem and a global manufacturing problem. He underscored that if the USAF wants to shorten time cycles, it has to improve trusted decision making in distributed environments by understanding each function and sharing information across boundaries.
He emphasized the value of moving from a domain approach for analysis to a survey of viewpoints (see Figure 2.2). Horizontal integration refers to the product, vertical integration represents the capability, and operations and logistics bridge the capability and the product. Humans would be supported by an enterprise integration with information-sharing technologies, advanced analytics, and other tools. A horizontal and vertical integration that enables a fit-for-purpose approach to the various viewpoints is also important, he continued. The viewpoints help address sociotechnical challenges across the product life cycle and enable the integration of domains (e.g., a design problem or a manufacturing problem becomes a product problem).
Dr. Hedberg pointed out that it is important to keep decisions as open as possible or to make a decision as late in the process as possible, because doing so expands the solution space. Making a poor decision owing to
a high level of uncertainty could add cost and inefficiencies to the rest of the life cycle. Part of this approach includes the ability to better buy down and accept risk (e.g., the commercial sector buys down risk every day). It is important to understand what capabilities are available and what they can do in terms of agility and flexibility, he continued. Logistics enable operational control to coordinate resources and keep the larger supply chain on schedule. The digital enterprise makes it possible to answer the following questions quickly using distributed data: (1) How well is the product performing to specification? (2) Where can machines be sacrificed today to ensure availability for the parts that could be delivered tomorrow? (3) What capabilities are available to produce parts? How are the capabilities performing to specification? (4) Are these processes performing the correct actions and using the appropriate capabilities to deliver the needed parts at the desired times?
Dr. Hedberg shared three opinions and suggestions:
- To win in tomorrow’s cyber and/or physical combat zones, the USAF must be able to predict where to engage and what is needed to engage, as well as rapidly understand and diagnose readiness to engage. The USAF could use a public–private partnership to develop a standards-based, flexible, and modular integration architecture for acquiring and delivering systems. Immediate actions to achieve this include leveraging technologies systematically by inserting and removing technology as needed in existing environments; studying and standardizing the interfaces (and understanding the inputs and outputs of domains); and increasing convergent, multidisciplinary project management office (PMO) teams that include configuration managers and logisticians (i.e., deploying model-based operational control).
- Policy is a larger barrier to success than technology. The digital thread and digital twin are unsustainable in the current environment (e.g., a digital thread related to the Next Generation Air Dominance Aircraft requires $80 billion to $180 billion for software development and sustainment; digital twin development and sustainment requires $1 trillion to $2 trillion). However, there are significant cost savings and efficiency gains across the defense industrial base with certain new technologies, and evidence to enabling an integrated
- smart-manufacturing approach could provide industry with a $100 billion annual savings opportunity. Current acquisition and sustainment policies extend the time cycle of deployments and fielding of systems, and DoD cannot outsource its technical and program management expertise and still expect to be able to obtain effective and affordable systems that can be delivered quickly. The USAF could focus more on relative-cost analysis than absolute cost; develop new ways to certify systems and assets that enable more model-based/simulation-driven testing, evaluation, verification, and validation; and resolve the great data rights debate.
- The USAF is best in class for fielding technologies, but adversaries are investing heavily and catching up quickly. The United States faces a blended attack (i.e., economic, intelligence, cyber, and physical) on key industrial sectors that support national defense and economic security. In collaboration with DoD and the United States, the USAF could holistically and systematically develop a strategy for rapid identification, protection, detection, response, deployment, and recovery of assets and capabilities in times of need or conflict. Immediate actions to achieve this include better integrating projects and PMOs with overlapping technology needs and investments; advocating for solutions to the “different pot of money” problems; and developing action forces that pull expertise from the research laboratories, federally funded research and development centers, and university-affiliated research centers (UARCs) to be rapid response teams to solve difficult problems quickly.
Dr. Hedberg noted that for 10–15 years, the USAF has been on a digital transformation path toward an extended enterprise using the concepts of MBE, digital thread, and digital twin. However, most advances have occurred in domain-specific siloes that do not advance the level of needed integration to allow for quickly designing, building, and fielding critical systems. Now, the USAF could develop a “whole of service” strategy that does not destroy the siloes but instead connects them with a digital thread, he continued. To maintain dominance, the USAF would benefit from open architectures and policies to support the “brownfield” realities of manufacturing and allow interconnectivity across decentralized systems. These distributed manufacturing systems would address this industrial need by coupling existing technologies with emerging “smart” technologies to enable advanced variant configuration—allowing the USAF to be agile and flexible enough to manufacture locally to the end user, in varying lot sizes, on-demand, with high first-time yields.
During the question-and-answer session, Ms. Westphal said that although war is a human endeavor, none of the 2019 NIST digital enterprise projects resemble human endeavors. She cautioned that in moving from domains to viewpoints, it is possible to reinforce the stovepipes that already exist. Viewpoints challenge current human biases and belief systems, which makes humans hesitant about change, so she wondered how all of these other viewpoints will be introduced. She noted that every time the USAF tries to maximize or integrate, the process grows increasingly complicated without consideration for the human component. Dr. Hedberg replied that one of the core competencies of the UARC is adding a human-in-the-loop. He added that the technology is straightforward; consideration for the human is more challenging. The digital engineering office focuses on technology, but human domain research (how humans integrate into these problems) is limited. Even with autonomy driving much of the progress, there will be humans (making policies or flying planes) who have to be integrated into the process. He agreed that it is critical to understand how that human interacts with the rest of the system. Ms. Westphal asked how to stop this cycle of focusing on technology more than the people, because humans (unlike machines) have the ability to trust. Dr. Hedberg concurred that the focus is often on the human–machine interface or the human–environment interface instead of the human–job interface and human–organization interface. He added that cost savings are not being realized by jamming new technologies into old processes, which is why it is crucial to have a holistic perspective. Integrated product development teams do not exist yet, and those types of problems will not be solved with technology, he continued.
Gen. Martin pointed out that the existing organizational relationships in the USAF are not set up to deal with a dynamic and integrated systems approach. He reiterated that there is no integrating structure for information management. Dr. Hedberg agreed and added that a matrix of convergent research teams could begin to look at problems from a human-domain perspective. Ms. Westphal wondered whether this approach focuses on the recognition of conflict, and Dr. Hedberg replied that the UARC is in the process of setting up proving grounds—for example, a cognitive security team is exploring “grey zone activities” and the new definition of warfare in terms of cyber.
AUTOMATION, THE DIGITAL REVOLUTION, AND THE IMPLICATIONS FOR TIME CYCLE REDUCTION
Dr. David A. Mindell, professor of aeronautics and astronautics, and Dibner Professor of the History of Engineering and Manufacturing, Massachusetts Institute of Technology; and founder and chairman, Humatics Corporation, shared an image of Alvin, a manned submersible developed by the Office of Naval Research in the 1960s that is still used by oceanographers to explore the ocean floor. He explained that the merits of human versus remote presence continue to be debated.
He described the remotely operated vehicle Jason, which is neither cheaper nor safer than Alvin as a means to explore the ocean floor, owing to the required cable with 30,000 lb of tension and 2,400 volts in salt water. However, Jason fundamentally changed the way that the sea floor was explored: It could stay down for 1 week, and it provided a “real-time seminar” for 40 people on the ship. Although Jason is classified as a “tele-robot” because it is controlled by a cable from the surface and brings data to the surface, it could accomplish little on its own in terms of intelligence or autonomy.
Dr. Mindell portrayed the move toward full autonomy via untethered vehicles as the next important phase of ocean floor exploration. Thousands of these vehicles are currently used by the U.S. Navy, for oil exploration, and for mapping. However, he explained that they are not truly autonomous because they are programmed by people, are sent by people down to the sea floor, and conduct pre-programmed track lines. The vehicles act as an extension of the human senses to digitize the sea floor by collecting and bringing data to the surface for people to analyze. These vehicles can be thought of as “tele-present” sea floor digitization tools, he continued. Untethered remotely operated underwater vehicles, such as Nereid, have the potential to disrupt the existing cost model for robotic intervention in a broad array of applications. He characterized autonomy as a method of time shifting, allowing a greater number of people more time to study the data; thus, autonomy is a function of bandwidth, time, and human context.
Dr. Mindell described his series of books that explore human interaction with machines in various types of systems. Between Human and Machine: Feedback, Control, and Computing before Cybernetics (2000) details early control systems and the analog revolution.6 He explained that naval gun control from the late 19th century through World War II relied on a type of human integrated system and autonomy. In the 1930s, while technology continued to mature, the antiaircraft problem emerged. The systems were unable to counter aircraft, which were much faster and presented a 3D problem. During World War II, the U.S. Navy reconfigured its research and development; universities became involved; and small, distributed solutions were created to address this antiaircraft problem. The Mark 14 Gun Sight was put on a 20 mm cannon that was operated by an individual sailor using two pneumatically controlled gyroscopes to calculate the lead. This device was very successful: 85,000 were produced during World War II. This exemplified the value of moving from centralized solutions to small, distributed solutions. Digital Apollo: Human and Machine in Spaceflight (2008) described computers and software built at Draper Laboratories for the lunar landing.7 The National Aeronautics and Space Administration asked Draper Laboratories to build a computer and a guidance system to land on the moon in 1960–1961. Believing they could land on the moon without people involved, the “interface” created included two buttons: “Go to Moon” and “Take Me Home.” The astronauts, however, desired a more collaborative approach. A human–machine system was created to guide the spacecraft to the moon, which enabled six successful landings. Lunar landings have to be considered as enabled by sociotechnical systems, Dr. Mindell commented. This is another example of operating mission-critical, real-time systems through bandwidth-limited, time-delayed telemetry links using autonomy as a way to unite the systems to accomplish the mission. In addition to his own publications, Dr. Mindell has spent several years advising USAF lieutenant colonels and colonels on their dissertations.8
___________________
6 D.A. Mindell, 2000, Between Human and Machine: Feedback, Control, and Computing before Cybernetics, Johns Hopkins University Press, Baltimore.
7 D.A. Mindell, 2008, Digital Apollo: Human and Machine in Spaceflight, Massachusetts Institute of Technology Press, Cambridge, Mass.
8 Example dissertations include The Sociotechnical Construction of Precision Bombing: A Study of Shared Control and Cognition by Humans, Machines, and Doctrine during WWII (Col. Raymond P. O’Mara); The MQ-9 Reaper Remotely Piloted Aircraft (RPA)t: Humans and Machines in Action (Lt. Col. Tim Cullen); and Tiger Check: Automating the U.S. Air Force Fighter Pilot in Air-to-Air Combat, 1950–1980 (Col. Steven A. Fino).
Dr. Mindell’s book Our Robots, Ourselves: Robotics and the Myths of Autonomy (2015) presents lessons learned from 40 years of remote and autonomous systems in extreme environments of sea, air, and space.9 He explained that systems in these environments become less autonomous as they are situated within human systems of operations and use—in other words, systems are successful only when they work closely with humans. Situated autonomy, he continued, is more desirable than full autonomy. A mix of human, remote, and autonomous capabilities leads to new modes of human presence. Because people will always be part of this endeavor, the following questions are important to consider: Where are the people? What are they doing? Are they acting in real time? Through what bandwidth and time delays? Why does it matter?
Dr. Mindell explained that telepresence is a form of automation; however, telepresence automates out many low-income jobs. He described his company, Humatics, which builds precision navigation systems to enable closer collaboration among people, machines, and infrastructure. For example, Humatics installed beacons in the New York City subway tunnels for train communication, enabling high precision, real-time knowledge of train locations. He indicated that the entire New York City subway system has essentially been turned into a robot that collaborates with 5 million people each day.
During the question-and-answer session, Ms. Westphal observed that DoD employs millions of people. She wondered how these employees would be utilized if automation technology comes to fruition in 10–20 years. Dr. Mindell hypothesized that fewer people would be exposed to the dangers of combat, and more people would be working in various support functions. This technology would not necessarily lead to a change in net employment, but it still has a substantial social implication for the structure of the military. Ms. Westphal asked specifically about impacts on the young people who are joining the military as a means to escape poverty. Dr. Mindell said that in his role on the Work of the Future Task Force, he discovered that the number of jobs overall has not decreased in the private sector, but technology has exacerbated the polarization of work. There are many jobs that are still difficult to automate (e.g., cleaning and service) at one end of the spectrum and highly paid knowledge jobs at the other. What has become hollowed out are the jobs in the middle of the spectrum, where career paths are increasingly difficult to find. He noted that the armed forces may have been more successful in maintaining career paths and middle layers of employment. The interim report from the Work of the Future Task Force was published in November 2019,10 and research briefs are published weekly.11 A final report12 will be issued in November 2020.
Dr. Daniel Hastings, department head, Department of Aeronautics and Astronautics, and professor of aeronautics and astronautics, Massachusetts Institute of Technology, referred to an ST Engineering robotics system in Singapore that replaced people who offload baggage from an airplane with technological solutions to turn the plane over faster. He asked if those types of jobs would shift instead of disappearing with increased use of automation technologies. Dr. Mindell remarked that it is likely that fewer people may be employed unloading an aircraft; however, if the cost of cargo decreases, people will be needed to move more kinds of cargo. He asserted that robots can enable a company to grow in a way that can increase employment, although the resulting jobs may not be “high-skilled” jobs.
Dr. Michael Yarymovych, president, Sarasota Space Associates, pointed out that driverless cars are not fully automated. Dr. Mindell replied that it could take at least another 10 years for driverless cars to be fully autonomous (if ever). He added that it is difficult to predict how long it will take for any technology to be adopted. Dr. William Powers, retired vice president of research, Ford Motor Company, asked if an automated car should ever function completely without a driver. Dr. Mindell highlighted that every technology that has emerged to benefit unmanned aircraft has also improved manned aircraft; he expected that to be the case for cars, too. Technologies such as brake by wire, steer by wire, and lane keeping are already in human-driven cars, and those technologies are likely to continue to improve. Even with “driverless” cars, it is likely that an operator will always be telling the car where
___________________
9 D.A. Mindell, 2015, Our Robots, Ourselves: Robotics and the Myths of Autonomy, Viking, New York.
10 D. Autor, D. Mindell, and E. Reynolds, 2019, The Work of the Future: Shaping Technology and Institutions, Massachusetts Institute of Technology, https://workofthefuture.mit.edu/wp-content/uploads/2020/08/WorkoftheFuture_Report_Shaping_Technology_and_Institutions.pdf.
11 For a list of these research briefs, see https://workofthefuture.mit.edu/research-type/briefs/, accessed December 29, 2020.
12 D. Autor, D. Mindell, and E. Reynolds, 2020, The Work of the Future: Building Better Jobs in an Age of Intelligent Machines, Massachusetts Institute of Technology, https://workofthefuture.mit.edu/wp-content/uploads/2020/12/2020-Final-Report3.pdf/.
to go, especially when the route needs to be shifted. It may not make sense to eliminate the role of the human completely, he continued. Dr. Powers followed up with a question about automating commercial aircraft and rail systems, and Dr. Mindell mentioned that there are “driverless” metros in Paris—a “driver” remains on the train but performs different tasks. The automation of aircraft presents a much more complicated problem that involves policy, regulation, and systems, in addition to the technology.
Dr. Engelbrecht asked if human–autonomous interface developments have compressed the amount of time in which operations occur. Dr. Mindell acknowledged that the speeds at which the vehicles operate compress the time; however, it remains to be seen whether those systems can be developed any faster. Wartime pressures often force people to solve problems and innovate quickly. Gen. Martin raised the issue of the ethical, moral, legal, environmental, and social implications of AI. Dr. Mindell commented that decisions are made by humans, and they may be made several years before an AI system is designed. Building a system includes programming in the view of the world that the human wants it to have. Much context goes into decision making, and it is difficult to build machines that can consider context. People have also become more aware of the bias and issues of integrity in the data drawn on by AI systems. He observed that AI is still very much a human product.
OPEN DISCUSSION
Lt. Gen. Wendy Masiello (USAF, ret.), president, Wendy Mas Consulting, LLC, noted that it is possible to broaden the talent pool if the need to centralize jobs in Washington, DC, is eliminated. Ms. Westphal mentioned that pre-COVID-19, there was speculation that 70–80 percent of the workforce would become freelance by 2025–2030. Lt. Gen. Bowlds pointed out that virtual work tools, such as Zoom, limit people to two-dimensional interactions and may not facilitate true collaboration. Lt. Gen. Masiello stressed the value of collaboration; without it, acceleration is not possible. Ms. Westphal suggested that instead of going fully virtual, money reserved for a physical building could be reallocated for employee travel to preserve crucial human–human interaction. She wondered if the USAF structure is too archaic—it is difficult to create a networked organization around hierarchy, pomp and circumstance, and promotion. Lt. Gen. Masiello added that having fewer people at the senior level could increase salaries across the board, which would enable both hiring the right people with the right talent and relying less often on external expertise.
Dr. Ryan highlighted the common misconception that conflict is a binary situation. Posturing, strategic communication, deception, and manipulation occur long before weapons are fired (during “peacetime”), which is important to consider when discussing reduced time cycles for conflict-related decisions. Gen. Martin underscored Dr. Hedberg’s guidance to delay decisions in order to keep the solution space open, which is the opposite perspective of that held by many senior leaders. This outlook complicates the question about the tempo at which the USAF should operate. Dr. Powers said that the time needed to make a decision depends on the type of decision. He suggested defining specific elements that hinder speed and discussing what could be done about each. Ms. Westphal noted that there are different time elements across the organization that interfere with increasing the speed of operations. Dr. Ryan added that speed is advantageous not only in the ability to drop an appropriate munition on a target but also for the time to build a talent pool: It takes approximately 10 years to train people for a particular capability. Reducing this time depends on the products of the educational system, she explained. Ms. Westphal observed that the ability to slow down the adversary is another aspect of speed. Dr. Richard Hallion, senior adviser, Science and Technology Policy Institute, emphasized that the ultimate goal is to improve efficiency across the USAF through the accrual of time, but a related goal is to drive the opponent to be reactive. Gen. Martin asked how the USAF uses the mentality that emerged in the 1970s to produce global battlespace awareness, precision, and penetration capability on its own terms instead of on the enemy’s terms. Ms. Westphal replied that synchronization and desynchronization are both important approaches to consider and emphasized that increasing speed is not the only way to win. She added that it is important to study issues beyond developing technology (i.e., organization, training, operation, and command structure). Dr. Ryan commented that the definition of “win” may also have to be revisited. Dr. Yarymovych asserted that the speed of protecting the nation is equally important to the speed of aggressing the opponent.
Ms. Westphal said that she would like to see “a small group of people with big brains” think about and begin to
work on these workshop series themes (e.g., challenging the status quo, direction, beliefs, and biases). Gen. Martin explained that the world has moved forward since the Cold War in terms of three vectors: (1) There are people and nations who are devoted to allowing the human spirit to thrive and its environment to grow. (2) There are other people and nations that want to control the human spirit (e.g., China, Russia). (3) There are people who believe in “God’s law” and that freedom and existence are subservient. He wondered how the United States prevents the latter ideologies from destroying its vision of liberating the human spirit and encouraging progress. A mechanism is needed to create progress, he continued, and tools exist that will be key for success. This relates to the ability to understand the nature of the battlespace from a global perspective and to be able to determine courses of action to stop, destroy, or deceive the enemy into thinking other things are happening faster than they can overcome them. Dr. Engelbrecht advocated for the creation of a framework that provides a description of time in the different dimensions of interest—for example, time needed to influence others, to educate the workforce and talent pool, to understand competitors, and to build relationships with allies. He added that there are both cultural and geographical dimensions as well as causes and inputs to time that lead to certain outcomes.
