Software and Robots
Adapting policies and practices around software development and autonomous systems is crucial for maintaining a military edge
Welcome back, defense enthusiasts.
This week, I have two compelling reports to share with you all. Their topics include:
The importance of software adaptability in modern defense strategy and its implementation by the DoD
How the US should pursue Confidence-Building Measures (CBMs) for autonomous systems in the Indo-Pacific
Quick Tanks is a weekly collection and summary of the latest long-form analytic content on the topics of US defense, force structure, innovation, and policy considerations. We strive to aggregate all of the key sources of analysis and present brief, neutral summaries to help keep you informed. Should you feel inclined to learn more about any study, please reference the full report via the links provided.
The sponsor of the newsletter is the Hudson Institute’s Center for Defense Concepts + Technology.
Tank you for sharing and subscribing, and happy reading.
Compiling Advantage
Unlocking the Competitive Power of Software Adaptability
By Ellen Lord and Dan Patt
Hudson Institute’s Center for Defense Concepts + Technology.
Link to PDF; Link to Report Page
Focus: This policy memo focuses on the critical importance of software adaptability in enabling the DoD to outpace competitors. It examines the challenges and roadblocks hindering the DoD's ability to achieve this vision and offers actionable recommendations for reform.Analysis: The memo draws insights from the authors' testimony before the House Armed Services Subcommittee on Cyber, IT, and Innovation and their extensive interactions with the technology and broader defense communities. It also references various DoD programs, initiatives, and case studies to support its arguments.Argument: The DoD must prioritize software adaptability to maintain its military advantage in an era of strategic competition. By embracing modern development practices, fostering collaboration with industry, and empowering software teams, the DoD can deliver capabilities at a faster pace and respond effectively to evolving threats.Insights: Units with the ability to adapt their software rapidly exhibit significantly higher operational efficiency compared to those relying on external updates. However, significant obstacles such as bureaucratic ATO (Authority to Operate) processes, lack of in-house technical expertise, and inflexible funding mechanisms hinder rapid software development and deployment within the DoD.Recommendations: See below for the full list of recommendations.This policy memo from Hudson Institute’s Center for Defense Concepts + Technology addresses a critical issue in the era of strategic competition among technologically advanced powers. The memo argues that the ability to rapidly develop, deploy, and update software is not merely an enabler of military capabilities but an emerging foundation of military advantage in the digital age.
“Once a conflict begins, adaptability and scaling drive outcomes. The Pentagon needs to seize the current moment to prepare. For an example of how conflict drives adaptation, consider that the lifecycle of a radio in Ukraine is only about three months. By then, radios typically need to be reprogrammed or swapped out as the Russians have optimized their electronic warfare against them. A new weapons system reaches peak efficiency about two weeks before countermeasures emerge.
For an example of a superior weapons system handicapped by unadaptable software, consider that Excalibur precision artillery rounds initially had a 70 percent efficiency rate when Ukraine first used them. After six weeks, Excalibur’s efficiency declined to only 6 percent as the Russians adapted their electronic warfare to counter the system. This shows how quickly adversaries can adjust to new technologies.
This lack of adaptability is not an inherent property of software; rather, it is a consequence of how we choose to manage software. Ukrainian units with organic programming capability that can rapidly adapt their unmanned aerial vehicle (UAV) software have about 50 percent efficiency, while units reliant on longer supply chains where feedback must be lowed back to an original manufacturer to make changes struggle to hit 20 percent efficiency. Keeping software in a pliant, fluid state is the only way to maintain tactical innovation.”
Challenges and Roadblocks to Software Adaptability within the DoD
Granting Software Authority to Operate: The current ATO process treats software as a boxed product, leading to lengthy reviews and documentation requirements that slow down the deployment of new capabilities and stifle innovation. This process has also created a new form of vendor lock-in, hindering the DoD's ability to tap into the latest innovations from across the commercial sector.
Resourcing: The DoD's budgeting and acquisition processes are largely geared toward traditional hardware programs with rigid requirements and long lead times, making it difficult for software development teams to secure the funding and support they need to iterate rapidly and deliver capabilities in a timely manner. Additionally, the inability to use a single appropriation to fund software improvements creates significant management challenges.
Talent Gaps: The DoD faces challenges in attracting and retaining top digital talent due to competition with the private sector for a limited pool of skilled professionals and the slow and complex nature of government hiring processes. Furthermore, acquisition personnel lack sufficient training in software acquisition pathways, various contracting methods, and the distinctions between software and hardware requirements.
Data: There is a lack of clarity within the DoD around key software concepts such as data rights, interface rights, and the appropriate role of industry in the software innovation process. Importantly, to train mission-oriented AI models effectively, software developers need both one-time and ongoing access to libraries of mission-relevant data.
Promising Developments and Best Practices
The introduction of the Adaptive Acquisition Framework (AAF) marks a significant transformation in the DoD’s procurement strategy. This new direction shifts from a universal approach to one that is more adaptable and bespoke, acknowledging the distinct needs of each acquisition project.
Additionally, the Navy's Program Executive Office for Integrated Warfare Systems (PEO IWS) has taken several groundbreaking steps. It has initiated a software factory, embarked on a continuous Authority to Operate (cATO) process, made systems interfaces accessible, adopted Modular Open Systems Architecture (MOSA) procurement methods, and spearheaded a portfolio management strategy for its extensive range of over 140 programs.
Moreover, the recent creation of the Chief Digital and Artificial Intelligence Officer (CDAO) position aims to dismantle data barriers, promote widespread access to system interfaces among program offices, and foster collaboration between government and industry partners to harness data for the development of powerful applications. However, while these initiatives indicate forward momentum, broad-based enhancements are necessary across the DoD to reach the intended level of software flexibility.
Recommendations
Enable personnel to rapidly deploy and update software: The DoD must focus on optimizing the software deployment workflow and overhauling the ATO procedure. This necessitates a transition from a rigid, compliance-oriented system to a dynamic, risk-managed strategy. It also involves promoting responsibility among authorizing officials, creating a uniform continuous ATO structure across the DoD, and advocating for ATO mutual acceptance across various DoD programs, services, and agencies.
Attract and empower top technical talent and foster a culture of doers: The DoD must enhance its appeal to and retention of premier technical professionals through an overhaul of its recruitment practices. This entails ensuring that new recruits enjoy autonomy in their positions, forging more adaptable career trajectories, and granting staff access to ongoing educational and development opportunities. The department ought to broaden the scope of current policies to facilitate temporary assignment programs, revamp the performance review process, and allocate resources toward the training and professional growth of acquisition personnel.
Prioritize APIs and data accessibility: The DoD needs to focus on making the APIs and interfaces of its current systems accessible and mandate that creators of forthcoming systems disclose their APIs and interfaces. This approach will enhance software flexibility, foster the creation of innovative strategies and operational methodologies, unleash data potential, and expedite the advancement of artificial intelligence technologies. The department should formulate explicit instructions and leading practices for API creation and oversight, release extensive data directories, inform the procurement and contracting personnel about MOSA, and collaborate with industrial allies to guarantee that essential interfaces are thoroughly documented, secure, and capable of expansion.
Embrace a diverse, software-centric industrial base: The DoD must strive to cultivate a broader, more software-focused industrial network capable of meeting its demands for flexible, creative software solutions. In situations where requirements are still too ambiguous for precise definition, the DoD should avoid overly detailed specifications. Instead, it should collaborate closely with its industry partners to progressively identify the balance between what is technically possible and what is needed for the mission, aiming to create valuable solutions.
Adopt recommended resourcing reforms: The DoD should enhance its existing resources and introduce novel procurement and financial mechanisms that reduce the timeframe required to initiate and green-light software endeavors, incorporating automated processes for reporting and scrutiny to improve supervisory practices. For instance, the concluding submission from the PPBE commission offers several suggestions conducive to more rapid software deployment. These include the flexibility to employ procurement, RDT&E, or O&M funds across the entire spectrum of software development, procurement, and maintenance stages.
I highly recommend reading the full policy memo to better understand the importance of software adaptability for the DoD and how the DoD can implement it.
Autonomy and International Stability
Confidence-Building Measures for Uncrewed Systems in the Indo-Pacific
By Tom Shugart
Center for a New American Security
Link to PDF; Link to Report Page
Focus: This report examines confidence-building measures (CBMs) for uncrewed air, surface, and undersea vehicles as a means to reduce risks of inadvertent escalation between the US and China in the Indo-Pacific.Analysis: The report uses qualitative analysis drawing on historical CBM case studies, existing US-China CBMs, international maritime/aviation agreements, and a tabletop exercise to test proposed CBMs.Argument: The proliferation of autonomous and uncrewed systems, coupled with the lack of experience in their operational use, necessitates the development of new CBMs to prevent accidental or inadvertent military escalations. The report posits that while CBMs have historically been challenging to implement effectively, they remain a critical tool for maintaining international stability, especially given the rapid technological advancements and the strategic rivalry in the Indo-Pacific region.Insights: CBMs allowing uncrewed systems as "half-rung" escalation tools could paradoxically enhance stability. Secondly, CBMs are more effective if focused on externally verifiable aspects of uncrewed systems. Moreover, unilateral US CBMs may have benefits even absent China's participation.Recommendations: US policymakers should make a unilateral declaration of CBMs for uncrewed systems, coordinated with allies and partners. It also calls for US and Chinese policymakers to negotiate "minimalist" bilateral CBMs. See below for more details.In the current strategic landscape, militaries worldwide are increasingly turning to autonomous and uncrewed systems to enhance capabilities, expedite deployment, and mitigate risks to personnel. However, the novelty of these technologies brings with it unprecedented risks. Within this context, this CNAS report presents a comprehensive examination of the necessity for CBMs, particularly in the Indo-Pacific region.
Confidence-Building Measures and Their Importance
CBMs are actions taken by states to reduce the risk of conflict and increase trust between parties. CBMs typically fall into four categories: exchanging information between parties, exchanging observers and/or conducting inspections, establishing "rules of the road" for military operations, and applying restraints on the operations and readiness of military forces. For example, the US-Soviet Incidents at Sea Agreement (INCSEA) of 1972 established a set of rules for safe navigation and communication between the two nations' naval forces, reducing the risk of accidents and misunderstandings at sea.
CBMs are particularly important in the context of autonomous systems, as these systems pose unique challenges such as the brittleness of algorithms, the potential for arms racing and threats to nuclear stability, and the risk that militaries might put too much trust in AI-driven systems. By establishing clear guidelines and protocols for the use of autonomous systems, CBMs can help mitigate these risks and promote international stability.
Building CBMs on Existing Agreements and Addressing Gaps
The report emphasizes the importance of building CBMs on existing agreements, such as the 2014 US-China MOU for Safety of Air and Maritime Encounters, the 2015 US-China MOU Supplement, the Code for Unplanned Encounters at Sea (CUES), and the International Regulations for Preventing Collisions at Sea (COLREGS). However, the report also identifies several gaps in these agreements that need to be addressed to ensure their effectiveness in the context of uncrewed systems.
Avenues for Establishing CBMs
The report proposes three possible avenues for establishing CBMs for autonomous systems:
Unilateral Declaration: The US could make a unilateral declaration extending the coverage of existing US-China agreements and CUES to US uncrewed vessels and aircraft. This approach would allow the US to establish new norms and pressure China, but it might also make China more suspicious of US intentions.
Minimalist Negotiated Approach: The US and China could negotiate minor additions or revisions to current agreements to explicitly extend coverage to uncrewed vessels and aircraft. This approach would be less controversial but may not address all the necessary gaps in the existing agreements.
Proactive Negotiated Approach: The US and China could negotiate more comprehensive additions and revisions to current agreements, including explicit clarifications for the expected behavior of uncrewed platforms. This approach would be the most thorough but may also be the most challenging to achieve given the current state of US-China relations.
As part of the report’s analysis, the author conducted a tabletop exercise in which teams played out scenarios with various levels of CBMs in place. Some important insights include:
“A potential stability-instability paradox exists for uncrewed systems, which could make them useful in a crisis. The proposed CBMs in the exercise were designed primarily to curb the chance of an accident or inadvertent escalation resulting from uncrewed system operations causing open conflict between the United States and PRC. However, as groups discussed the utility of the CBMs in the scenario, they stressed that uncrewed systems also might be useful to help defuse a crisis from escalating into violence. In particular, the reduced emotional impact of losing an uncrewed system versus human lives could act as a pressure-release valve for the countries looking to coerce the other side during a crisis without escalating to war. Participants stressed that uncrewed systems were special and different than crewed assets precisely because they allowed decision-makers to mitigate risk in otherwise dangerous scenarios. This created a stability-instability paradox where there might be significant contestation—and even bloodless violence—because of the increase in uncrewed systems, but that the use of machines over human platforms also may create a kind of stability when it comes to violent escalation, allowing states to de-escalate even with the loss of valuable uncrewed systems.
Uncrewed system CBMs are potentially most useful before and after a crisis, but not in the midst of one. Most of the groups believed that the crisis was both too heightened and at the same time not significant enough for CBMs to play a large role in the outcome of the crisis. This led players to debate when CBMs might have the largest role in avoiding accidents or inadvertent conflict. The group concluded that there are potentially two times when CBMs are most effective: (1) During normal, but perhaps slightly heightened tensions; for example, during a U.S. freedom of navigation operation or an isolated show of force or resolve; and (2) post-crisis, if a failure of crisis control leads both sides to conclude CBMs are in their best interests.”
Recommendations for Establishing CBMs
Ultimately, the report recommends a three-step process for establishing CBMs for uncrewed systems in the Indo-Pacific:
“First, adopt a unilateral declaration of uncrewed CBMs, announced in coordination with plurilateral CUES-specific measures by other like-minded allied and partner CUES signatories (recommended focus: Australia, Canada, France, Indonesia, Japan, New Zealand, the Philippines, Singapore, and South Korea). Such an announcement would include a statement of the right, at any time, to suspend these measures and revert to the existing CBM framework.
Next, negotiate through existing U.S.-China maritime consultative mechanisms “minimalist” measures, including the right of both sides to suspend such measures.
Finally, pursue more comprehensive measures including:
1. Providing working frequencies for uncrewed aircraft and vessels allowing for communication with crewed platforms, as feasible
2. Avoiding blinding uncrewed platforms’ navigation and collision-avoidance sensors, and control and communication systems
3. Agreeing that both sides’ uncrewed aircraft should operate in a manner that minimizes hazards to other aircraft
4. Agreeing that neither side will use uncrewed ships, aircraft, or undersea vessels to carry, deploy, or employ nuclear weapons
5. Agreeing that either or both sides, at any time, may declare an exemption of uncrewed systems from these measures for as long as required—except for measures related to nuclear weapons.”
To better understand how CBMs could play a role in the Indo-Pacific, I urge you to read the full report.