Digital Transformation and the DoD

Introduction

Prior to attending Harvard Business School, I travelled the world as an Officer in the United States Army.  My military deployments took me to faraway lands and exposed me to new peoples and unfamiliar cultures.  In 2016, I continued my journey, this time reaching the world of business by way of Harvard.  Over the last two years, I was again exposed to new perspectives and new ideas, but I never forgot that I would be returning to the military after graduation.

During many a class discussion, I found myself pondering the potential application of technologies or business frameworks in the context of the military.  Could ideas from the business world be leveraged to improve the lumbering, bureaucratic Department of Defense (DoD)?  As the rate of digital innovation accelerates into the future, the DoD, like many large, incumbent corporations, will be faced with many challenges, as well as significant opportunities.  How can the Department of Defense embrace digital transformation as a means of sustaining the competitive advantage needed to deter aggression and preserve the rules based international order?

Current Challenges

The DoD’s challenges are many and include budget constraints, foreign adversaries, human capital recruiting and retention, and technology acquisition.  As in business, digital transformation offers the DoD many opportunities to address its most pressing challenges with new technology.  In 2014, the DoD announced what is known as its Third Offset Strategy, “an effort to focus [its] innovation efforts on preserving and revitalizing its conventional deterrence capability by adapting to the countermeasures to key U.S. capabilities that near-peer competitors have built up in recent years and are continuing to develop.”[i]   But despite the DoD’s digital aspirations, digital transformation has become a challenge of its own for several internal and external reasons, including the following:

Internal

• Lack of Awareness and Understanding. The recent testimony of Facebook’s CEO before Congress revealed an alarming fact:  Congress does not understand the new digital world.[ii]  Congress controls the budget and sets the laws as they pertain to the DoD and if they, as well as senior military and civilian leaders within the DoD, are unaware of the opportunities of digital, they will be also oblivious to the changes that digital transformation requires.  While there are pockets of understanding and advocates, there is not an overwhelming sense of urgency.
• Resources and Talent. The DoD does not currently have the right talent to execute a digital transformation.  The DoD lacks engineers, managers, and leaders with these skills and talent and must therefore hire them.  The DoD struggles to fill this skills gap because it cannot pay competitively and has a relatively negative reputation in the tech community.  Also, because the DoD is a federal employer, it cannot easily fire people who resist digital initiatives.  Finally, despite its $500+ billion annual budget, the DoD has not dedicated resources to digital transformation.  Resources are divided amongst traditional subsections, with no unifying program or initiative.
• Buying cutting edge technology is hindered by significant regulation.  The DoD’s acquisition system has been described as “an 18th-century wooden warship that has been out to sea for too long, accumulating such a surfeit of barnacles that it can barely float, let alone operate under full speed.”[iii]  Those barnacles represent decades of reforms and regulations targeting previous mistakes, contracting abuses, or failed programs.  Instead of improving outcomes, the regulations have slowed the process.
• Process Design. The DoD acquisition process attempts to optimize solutions across diverse minimize risk, but it is designed for hardware, not software.  Today, software half-life is less than the time it takes to procure it, often changing three times over between when a contract was signed and the hardware fielded.[iv]  “With culture and process among the many obstacles, going from napkin sketch to battlefield deployment in two and a half years was the equivalent of traveling at light speed.”[v]
• Culture.  The DoD is not a uniform organization with a standard culture, but several common themes can be found to be at odds with a digital transformation: (1) the DoD is and should be the source of innovation in all military innovations, (2) aversion to risk and failure, and (3) pursue process over outcome (you won’t get fired if you follow the process set by law).

External

• Pace of Change. Technology is changing faster than ever before, and digital innovation will likely continue to accelerate into the future.  The impacts of these changes will reach every aspect of human existence, to include international competition and warfare.  For thousands of years wars were fought in only two domains: land and sea.  Within the last 100 years, the domain of air was developed, with its fighter jets and drones, and now in the last decade, technology has added two more domains of conflict: cyber and space.  The pace of this change is faster than the DoD’s current ability to adapt.
• International Competition. The implications of the United States’ adversaries to leverage digital technologies is “profound and broad” and include an “increased vulnerability to cyberattack, facilitation of advances in foreign weapon and intelligence systems, the risk of accidents and related liability issues, and unemployment.”[vi]  Indirect competition in digital also exists, as countries race to attract talent and invest in startups ahead of others.
• Unwilling Partners. In Silicon Valley, the DoD, and government more broadly, has a reputation for moving slow.   Venture capitalists often steer their portfolio companies away from DoD, seeing government as a risky customer with a longer time horizon to exit.  Startups also avoid DoD due in part to a lack of familiarity with the DoD’s problems and their inability to financially survive the DoD’s long acquisitions process.  Even within larger, technology companies, there is some resistance to working with the DoD.  For example, over 3,100 employees at Google signed a letter to the CEO, protesting their work on a DoD imagery analytics project.[vii]
• Ethical Concerns. As the DoD looks to digital transformation, many domestic and international voices have raised concerns.  Some question how the technology will be used.  Other activists have concerns over data privacy, would the DoD illegally obtain data or use data to target U.S. Citizens?  Overall, will the DoD use advances in technology responsibly in both peace and war?

Opportunities

As alluded to earlier, digital transformation is not a panacea to the problems of the DoD, but it is an essential tool.  Potential applications of digital transformation within the DoD could include:

Data

The DoD is awash with data and is passing on thousands of opportunities to capture other data and utilize it.

• Personnel recruiting. New recruits currently take a written test, designed in the 1960’s, known as the ASVAB to determine their suitability for military service.  The DoD could utilize hundreds of features and millions of data points from successful service members to create algorithms that use more than just a written math and English test to select the best recruits and pass on others.  Companies like Pyometrics already do this for big companies, why can’t the DoD?
• Equipment maintenance and operation. The DoD owns aircraft carriers, airplanes, tanks, and trucks and spends billions of dollars each year to repair or replace them.  Readiness for missions is hindered when equipment is inoperable due to maintenance.  Companies like General Electric sell data driven solutions, such as jet engine up time.  Why can’t the DoD purchase fighter jet up time or use sensors to monitor and predict tank engine failure?
• Threat pattern analysis and prediction. Everyday the DoD is under attack.  Whether it is a roadside bomb in Afghanistan or a cyber attack on the Pentagon, the DoD responds to and records the data from thousands of attacks each year.  Companies like Palantir already sell software solutions to DoD for some of these problems, but has every opportunity been exploited?

AI

Advances in AI will continue to disrupt businesses and can also be applied within the DoD, both on and off the battlefield.  In 2014, Vladimir Putin commented that “artificial Intelligence is the future, not only for Russia, but for all humankind.  It comes with colossal opportunities, but also threats that are difficult to predict. Whoever becomes the leader in this sphere will become the ruler of the world.”[viii]  Some have called AI, the next arms race, and the DoD should be very concerned if the United States is not winning that race. 

• Narrow AI for administrative tasks. Narrow AI can be used to optimize administrative tasks such as job matching, pay, medical records, travel vouchers, customer service, and insurance.  These tasks are currently paper based, repetitive, and slow due to limited human resources, but narrow AI could increase speed and user satisfaction and lead to cost savings.
• Narrow AI for intelligence analysis. The DoD collects millions of data points of intelligence every day, but does not have enough human analysts to evaluate all of them.  Narrow AI can tell if a picture is of a dog or make style recommendations for clothing, so why can’t narrow AI be used to interpret satellite imagery or other forms of intelligence.  Such a capability could augment human analysis, but should not replace it.
• General AI for decision making. General AI, though not a reality today, is becoming more capable each day.  For example, DeepMind AI’s AlphaGo Zero taught itself to play the complex, strategy game of Go, beating 60 top professional players, including the world champion.[ix]     Decision making today, at all levels of the DoD, is limited by the capacity of the human mind to sift through data and make recommendations.  These processes, such as the Military Decision Making Process, are subject to human flaws such as group think and confirmation bias.  General AI for the military should not replace human decision makers but could be used to augment and inform human decisions.
• General AI for aircraft and tanks: As with decision making, aircraft and other military technologies are limited by human abilities, skills, and experiences. For example, engineers for the Air Force’s new F35 joint strike fighter, had to limit the aircraft’s performance to a level that pilots could handle.  What if a fighter jet with General AI was trained with every flight simulation scenario and every record of air battles from the last 100 years?  Could it make split second maneuver decisions faster than human pilots, increasing pilot survivability?

Other areas of digital opportunity for the DoD include crowds and platforms.  Platforms inside the DoD could connect end users with problems to program officers with R&D budgets, connect leaders with similar problems, and share knowledge.  Crowdsourcing could be used to gather human intelligence or source solutions to capability needs.

Recommended Changes

Much of the technology listed above already exists today and the DoD simply needs to buy it or partner with industry to modify it to DoD needs.  In some cases, the DoD is already pursuing exploring similar opportunities.  In others, the DoD needs to rethink the way it does business.  The DoD cannot simply digitize the current means of operation.  Rather, the DoD must totally redesign or build anew if it wants to take full advantage of the digital transformation.

In 2014, the Federal Government launched the U.S. Digital Service, a federal consulting service in the Executive office of the President of the United States, that partners “leading technologists with dedicated public servants to improve the usability and reliability of our government’s most important digital services.”[x]  Since its launch, the program has expanded to other parts of the federal government, to include the DoD, but small consulting teams of outsider talent will not be enough to transform the DoD.  To accelerate the DoD’s digital transformation, it should consider the following:

• Educate.  Educate senior leaders to foster awareness and create an impetus for change.   Senior leaders must take a deliberate interest in and should personally shepherd digital initiatives to protect them from the bureaucracy.  Education should also expose mid-level managers, including career DoD civilians, to create awareness of what is possible.
• Create autonomous teams. Create small, independent teams with directives to solve certain DoD problems with digital solutions.  Teams should not be limited by traditional functional siloes within the DoD and should be encouraged to look outside the DoD.  Teams should also be exempted from DoD constraining regulations.
• Team with early adopters. Special Operations Forces tend to be outside the box thinkers and are traditionally early adopters within the DoD.  The autonomous teams should partner with SOF to test certain digital solutions, before rolling out to the rest of DoD.  Typically, when SOF has something new that works, the rest of DoD starts to ask for it.  Other units and offices within the DoD should be identified as test subjects and be the focus of education and prototype equipment and service testing.
• Incentivize.  Create promotion incentives that reward digital transformation and do not punish the “fail fast” mentality.
• Achieve quick wins. Pick visible, easy to solve problems and provide quick wins that demonstrate the value of digital transformation.
• Prioritize with dedicated resources. The DoD is known to label many things as a “priority,” but does not follow up with the resources due to the overwhelming number of priorities.  If the DoD is serious about achieving a digital transformation, it needs to be a resourced priority.
• Relationships.  Increase engagement with the tech industry.  Educate the ecosystem about the DoD’s problems as well as its commitment to responsible implementation.
• Regulations.  Find advocates within Congress and build a coalition to drive regulatory change.

Resistance to Change

Resistance to digital transformation within the DoD is an ongoing challenge, for many of the reasons highlighted above.  Two areas of resistance that the DoD should address include:

• Ethical Resistance. The DoD should engage in the national conversation instead of avoiding it.  By doing so, the DoD will increase transparency, engender trust, and increase its ability to influence regulatory outcomes through a more balanced national debate.  With respect to AI, the DoD has already pledged its commitment to keep a “human in the loop,” but such policy stances are not well known across the tech community.
• DoD Employee Resistance. Last year, I spoke with the former Secretary of the Navy, Ray Mabus and he related a story in which a legacy civilian employee working in the basement of the Pentagon told him he could not start a certain initiative despite the fact that he was acting on direct orders from from President Obama.  Despite best efforts, the culture, the people, and the regulation in some parts of DoD will be too resistive to digital transformation.   To address this resistance, the DoD should focus its efforts on those tribes who are traditionally early adopters.  In other cases, it should build digital capabilities in parallel to resistive, analog entities and quantitatively demonstrate the superiority of the digital solution.

Conclusion

While digital efforts within the DoD, like the Defense Digital Service, are underway, they are small and alone will not be enough to overcome the obstacles to digital transformation.  U.S. Secretary of Defense James Mattis recently expressed, “I’m certainly questioning my original premise that the fundamental nature of war will not change. [Because of AI] you’ve got to question that now. I just don’t have the answers yet.”[xi]  James Mattis is aware, but is it enough to jump start the transformation the DoD needs to sustain its competitive advantage?

Endnotes

[i] https://csis-prod.s3.amazonaws.com/s3fspublic/publication/170302_Ellman_ThirdOffsetStrategySummary_Web.pdf?EXO1GwjFU22_Bkd5A.nx.fJXTKRDKbVR

[ii] https://qz.com/1251569/if-the-us-congress-cant-understand-facebook-were-in-trouble/

[iii] https://www.technologyreview.com/s/603084/the-pentagons-innovation-experiment/

[iv] Ibid.

[v] https://www.fastcompany.com/40401930/how-military-veterans-turned-entrepreneurs-are-disrupting-the-pentagons-weapons-program

[vi] https://www.zdnet.com/article/ai-iot-and-the-end-of-moores-law-add-to-us-national-security-worries/

[vii]https://www.nytimes.com/2018/04/04/technology/google-letter-ceo-pentagon-project.html?mtrref=www.google.com&gwh=8107045A1831AC0ECB65A9618649CB75&gwt=pay

[viii] https://www.cnet.com/news/elon-musk-we-are-summoning-the-demon-with-artificial-intelligence/

[ix] https://deepmind.com/blog/alphago-zero-learning-scratch/

[x] https://www.usds.gov/

[xi] https://www.c4isrnet.com/intel-geoint/2018/02/17/ai-makes-mattis-question-fundamental-beliefs-about-war/