As the United States remains on a path towards continued reductions of nuclear weapons in concert with Russia, there is a likelihood that future arms control initiatives may include individual warheads – strategic and tactical, deployed and non-deployed. Verification of such an agreement could prove to be challenging and costly under an inspection-oriented regime such as that employed by the New START Treaty. As such, the concept of actively monitoring warheads throughout their lifecycle is proposed as a potential solution. An active monitoring system could reduce the burden of inspection activities to achieve equivalent confidence that treaty obligations are being upheld by increasing transparency of operations. Concerns about the sensitivity of data generated are warranted, and generating sufficient trust in the validity of data produced by this system is challenging, yet they are not insurmountable with a thoughtful design. This article explores the active monitoring concept, in addition to highlighting both the challenges and solutions such a system would provide.
The Obama administration has clearly stated an interest in continuing reductions of the United States nuclear weapon stockpile in accordance with Russia and the other nuclear weapons states. The New START treaty 1, signed in 2010 and ratified in 2011, limits strategic deployed warheads to 1,550 on 700 deployed delivery vehicles, with a total limit of 800 deployed and non-deployed delivery vehicles. In a 2009 speech in Prague, prior to the New START negotiations, President Obama brought a new focus to nuclear arms control by affirming “… America’s commitment to seek the peace and security of a world without nuclear weapons.” 2 While also admitting that this is very much a long-term goal, this statement and others made in the same speech set the policy of the United States as seeking to advance arms control goals beyond New START. After stating his plan to negotiate New START, he said that “… this will set the stage for further cuts, and we will seek to include all nuclear weapons states in this endeavor.” Further cuts may happen in a similar fashion to the START and New START treaties – reductions in the numbers of strategic, deployed delivery vehicles and warheads – though as those numbers continue to drop, the numbers of non-deployed and non-strategic (tactical) weapon systems and warheads become more prominent in the debate.
According to the 2010 Nuclear Posture Review, “… the Administration will pursue discussions with Russia for further reductions and transparency, which could be pursued through formal agreements and/or parallel voluntary measures. These follow-on reductions should be broader in scope than previous bilateral agreements, addressing all the nuclear weapons of the two countries …” 3 Under New START, all strategic delivery vehicles (missiles, land-based launch tubes, submarine launch tubes, and bombers) are accountable and limited, whether they are deployed or not. A follow-on agreement to New START that limited nuclear warheads and bombs (whether they are deployed or not), would shift the focus from accounting for the delivery system to accounting for the warhead, whether it is mated to a delivery vehicle or not. In addition, an agreement that limited non-strategic warheads and delivery systems would increase the scope of limitations: mildly for the United States, and significantly for Russia.
The shift in focus from delivery systems to warheads and the inclusion of non-strategic systems will make verification of the treaty terms much more difficult. In general, strategic systems are much easier to see from a distance than non-strategic systems and especially individual warheads. In addition, the set of locations that warrants inspections when including non-strategic systems and warheads (in storage, maintenance, etc.) is much larger than the set of locations under New START. Increasing the scope and number of on-site inspections to account for all nuclear weapons may not be desirable due to the large expense to the inspecting nation and impact to operations of the host nation. Therefore, new technical approaches for verification could become useful to ensure that arms control agreements will be maintained and trusted when the scope extends to all nuclear weapons – deployed and non-deployed, strategic and non-strategic.
The Verification Challenge
The verification methods used for New START are essentially the same as those used under START: (1) national technical means, (2) data exchanges and notifications, and (3) on-site inspections. 4 National technical means includes all manner of viewing and sensing the actions of the treaty partner from a distance, relying on national intelligence capabilities. Data exchanges and notifications are declaratory tools used to communicate the numbers and locations of all treaty-accountable items (TAIs) at the beginning of the treaty enforcement, at periodic intervals, and when things change. On-site inspections are used to verify those declarations by sending an in person delegation to a limited number of sites in the treaty partner country to view the TAIs at that site. There are two types of New START on-site inspections: Type One inspections focus on sites with deployed and non-deployed strategic systems, while Type Two inspections focus on sites with only non-deployed strategic systems (sites without warheads). During Type One inspections, inspectors have the opportunity to count the number of deployed strategic delivery systems and verify for a single delivery system (including a bomber at an air base), the number of warheads emplaced on it. The relevant inspections for this discussion are Type One.
The goal of verification is to generate a sufficient amount of confidence that the treaty partner is fulfilling their obligations expressed in the treaty. With effective national technical means, fewer and less intrusive on-site inspections are necessary to gain sufficient confidence. When the focus of reductions, and therefore of verification, shifts from strategic delivery systems to warheads and non-strategic systems, national technical means will be less effective. This result could mean that with more intrusive on-site inspections (and probably more inspections with the expanded set of locations of interest), the same amount of confidence can be generated in a new treaty as is generated by New START verification. However, with more inspections that are increasingly intrusive, costs for both sides rise and the impact to host operations suffers, since operations will likely be suspended at the site being inspected for the duration of the inspection.
Passive tags and seals have been suggested as assisting in verification of warheads: a warhead in a container could be sealed, and if the inspector verifies a seal on inspection the inspecting party has some confidence in the integrity of that particular warhead going back to the time it was sealed. But passive seals can only indicate that a seal was broken or not broken. No additional information about a broken seal is available, such as when or why the seal was broken.
An alternative and more comprehensive approach is to use active tags and seals, along with fixed monitoring devices in facilities of interest to create trustable information about the location and integrity of all TAIs. An active monitoring system in support of a future arms control agreement that includes all warheads – strategic and non-strategic, deployed and non-deployed – could reduce the cost of generating sufficient confidence enough to make the agreement feasible, while providing an unprecedented level of transparency.
Active Monitoring Approach
In lieu of increasing inspection frequency and complexity, an active monitoring system could be used to generate sufficient confidence that treaty declarations are being upheld while lessening the burden associated with inspection costs and the impact on operations at military installations. The approach of active monitoring discussed here uses an active tag with a monitored seal, known as an item monitor, which communicates to a centralized data collection point. After being attached and sealed to a TAI, the item monitor and associated data management system provides an indication of where the TAI is at any given point within the nuclear security enterprise –in storage, staging, maintenance, transportation, or deployment. The seal is designed to monitor when the item is physically removed from its handling gear which can occur during shipment, maintenance, or when deployed on a delivery vehicle. The seal design precludes removal of the warhead from its handling gear without breaking the seal. Additional layers of monitoring such as motion detectors, cameras, and other sensors can be added into the system to gather supplemental data and improve transparency of operations, while providing greater confidence in the information generated by the item monitors.
While all nuclear weapons in each country would be accountable and thus part of the monitoring regime, each TAI might not be actively monitored in every stage of its lifecycle. Figure 1 illustrates the seven generic stages of nuclear weapons in the United States, along with the dispositioning stage, which may be of interest for monitoring to account for latent nuclear weapons beyond dismantlement.
The deployment stage shown in the figure specifically represents warheads deployed on a delivery vehicle, and not those in storage at a deployed base (which are still considered in the storage stage).
Refurbishment of a weapon occurs as part of a Life Extension Program in which many components are replaced, whereas weapon maintenance implies a less significant replacement or access to the weapon without replacement, which can be done at the deployment or storage location.
Staging indicates that a weapon is awaiting refurbishment or dismantlement.
Dispositioning is the stage in which the dismantled weapon components are rendered unusable without an effort equal to production of those components.
Figure 1: Generic Nuclear Weapon Lifecycle Stages
As indicated in Figure 1, TAIs in the staging and storage stages would be continuously and actively monitored. Any integrity breach or movement during these stages would be recorded by the system. The transitions from the production 5 and to the dismantlement stages, as well as the transitions to and from the refurbishment, maintenance, and deployment stages would be recorded, though once the TAI is in any of those stages it would not be actively monitored.
Using the United States as a model, there are numerous sites where an active monitoring system would be installed to meet the requirements of a future arms control monitoring regime. Furthermore, within each individual site there could be multiple holding locations for weapons. At each site the information from each holding location would be aggregated and transmitted to a site-wide database. The information from the nation’s weapon sites would then be aggregated at the national level, reviewed, and periodically transferred to the treaty partner who would analyze it to verify declarations as well as discover undeclared activity. Thus, the concept of data exchanges and notifications currently used for New START verification would be retained, albeit with much larger sets of data and potentially more frequent notifications. The treaty partner could then select a sampling of locations and TAIs to inspect to increase confidence and ensure proper system functionality. The concept of on-site inspections would also be retained from New START, though the active monitoring system would limit the number needed to achieve sufficient confidence. A simplified view of this system is shown below in Figure 2 for three separate sites, each with three discrete TAI locations (either storage or maintenance).
Figure 2: National Model of an Active Monitoring System
In Figure 2, looking at a particular site there is a single TAI that is sealed and tagged by an item monitor moving from a storage area to a maintenance area and back to a different storage area. While it is in the maintenance area, the seal is broken and the item monitor is removed so that the warhead can be accessed for maintenance. Following the work, the warhead is placed back in its handling gear, which is sealed once again. In each of these locations the item monitor communicates with a data collection unit in the room, sending information during entrance and exit, as well as periodically throughout its existence in the room. In addition, fixed monitoring nodes in each of these locations (such as door switches, motion detectors, and cameras) generate additional information to create layers of evidence. The information generated by the monitoring system in each location – by item monitors as well as fixed monitoring nodes – is passed to a central data aggregation point at the site that combines the information from all locations at the particular site. Each site then passes information to a national data aggregation point, which is then transferred to the treaty partner during periodic data exchanges and more frequently during notifications.
All nuclear weapons that are properly maintained will still require routine maintenance and refurbishment, and these activities will likely occur without inspectors present to avoid releasing weapon design information. In order for the monitoring system to increase the treaty partner’s confidence in the host nation’s declarations of TAI activity, they must first trust that the TAI being monitored is an authentic nuclear weapon – i.e., that the host nation is not playing a shell game. As shown in Figure 3, at the start of a future agreement all TAIs would need to be verified as authentic in what is considered a baseline inspection, and then sealed using the item monitor while the inspecting partner is present. This baseline inspection likely would include measurements of attributes that are agreed upon in negotiations.
Following the baseline inspection at all sites, every nuclear weapon would be entered into the monitoring regime. A TAI with an item monitor attached (and sealed) goes from black to white. In the white (sealed) state, the treaty partner has confidence that that particular TAI is authentic, and thus trusts the information the TAI generated by the monitoring system. The TAI would then continue to move throughout the nuclear security enterprise as required by the host country, with its movements and the status of its seal being continuously monitored. Since nuclear weapons are not static items for the life of a treaty, seals will have to be broken and most likely TAIs will have to be removed from active monitoring for maintenance, refurbishment, and deployment. When performing a maintenance activity on a sealed warhead or preparing a warhead for deployment, the activity would be declared in the same dataset that is transmitted to the treaty partner. Normal operations would not require the presence of an inspector.
Once declared, the seal can be removed and the warhead operation can proceed. After the seal has been opened on a TAI, the authenticity of that item cannot be confirmed until it is inspected by the partner nation, which would likely include the same type of measurements made during a baseline inspection. At that point, the combination of re-establishing the authenticity of the TAI with the record of the TAI being sealed back to a point in the past gives the treaty partner confidence in the TAI from the time of sealing (indicated by the cross-hatched TAI in the figure), even if the treaty partner did not witness that sealing.
Figure 3: Timeline of Trust for a TAI
A monitoring system that accounts for individual weapons under a new arms control regime must have two basic characteristics: reliability and trustworthiness. Reliability implies that the system will work as intended with little or no downtime and without generating false information. While reliability is an important attribute of any engineered system, it is especially important in an arms control monitoring system. Any unexpected system behavior or relatively long downtime is likely to raise suspicion in the treaty partner, and would likely require a host country explanation. Trustworthiness is more complex. A system can be trusted by the host if the individual components and software can be shown to not interfere with the safety, security, and reliability of the nuclear weapons or the facilities that house the nuclear weapons (the process of certification). The system can be trusted by the treaty partner if the data it generates can be authenticated, it is hard (i.e. expensive) to forge false data, and the hardware and software used can be verified to not have hidden functionality (the process of authentication). Hardware and software authentication is challenging due to the complexity of integrated circuits and modern programming languages. Authentication concerns could be eased through either a jointly designed system or random sampling of the active monitoring system’s components by the treaty partner to inspect, possibly destructively. Data authentication requires the ability to digitally sign and verify the signature of the data generated by individual item monitors and fixed monitoring nodes, which necessitates the use of cryptographic algorithms to greatly increase the difficulty in forging messages. The system must also take into consideration the usability of the data from the perspectives of both the host and treaty partner to ensure that it is easy to sort and analyze the large quantity of data that will inevitably be collected.
The extent to which each side will assess the system equipment during certification and authentication also depends on who designs and produces the equipment. With host-designed and produced equipment, certification will likely be easier but authentication may be harder. With inspector-designed and produced equipment, authentication will be easier, but certification will be much harder, maybe impossible. A third option (which needs more study), is joint design and third-party (monitored) production. For our analysis, we have assumed host-designed and produced equipment.
The level of transparency associated with the active monitoring approach described here goes beyond any previous sharing of information under former treaties and agreements. Achieving concurrence and buy-in from stakeholders will be challenging – particularly the military services whose base operations may be affected – though the impact of a monitoring system may be less than the impact of the number of on-site inspections necessary in its absence. Additionally, many sensitive and potentially classified characteristics of the nuclear security enterprise could be revealed through the data aggregation and analysis process. To maintain the high level of transparency required for such an arms control regime, it may be necessary to redact portions of the data prior to transmitting it to the partner country. This could be done without degrading the integrity of the remaining data, but still providing enough information to account for warheads in the regime.
Potential arms control initiatives that include limits on total nuclear warhead stockpiles (including non-strategic and non-deployed weapons) and monitoring of warheads awaiting dismantlement may require technical accountability measures that are distinct from the technical measures used in previous treaties. Accountability measures could include active monitoring systems that provide trustable information and assurances of the location and the integrity of nuclear weapons and its components throughout the nuclear weapons lifecycle. Better understanding of active monitoring capability options for declared warheads and potential operational impacts of such a monitoring regime will help prepare for possible future initiatives.
Many challenges to the development and use of a nation-wide monitoring system in the U.S. and its treaty partners in support of a future arms control initiative remain. The scope of technology necessary for this system is much larger than what is used today for New START verification. The sheer complexity will make negotiations long and challenging. Generating trust with this technology may not be easy. Trustable components and information will be a key system attribute to be factored into design. The inspectors must trust the system to generate authentic and correct information, and to be highly resistant to undetected tampering by the host party. In addition, the host must accept the use of this equipment on or near nuclear weapons in their custody, which requires mitigation of concerns about safety, security, and divulging sensitive information. Lastly, no matter how well designed the system, on-site inspections would still be required to verify that the data generated by these systems reflects reality. However, the number of inspections could be minimized while still creating a level of confidence that is statistically significant.
Active monitoring of all nuclear weapons by a system coordinated across all staging, storage, maintenance, and deployment sites may be a key step in building confidence in such an agreement and reducing the need for on-site inspections to the point where the agreement is realizable. While 100% confidence in verification will be difficult, a system can be engineered to increase confidence that an agreement is being upheld by identifying the location and status of each TAI in an assured and trusted way to the monitoring partner, as well as providing layers of evidence of monitoring activities using various sensors and imagers. A flexible system will allow weapons to be accounted for and actively monitored through various phases of their lifecycle, thus enabling verification and increased confidence in weapons reductions. Research into the concept of an active monitoring system, including the operational impacts of such a system – and technology to support the concept – should be an element of a research agenda to support future negotiations for a new bilateral or multilateral arms control agreement.
Jay Kristoffer Brotz is a Senior Systems Engineer in the Nuclear Monitoring and Transparency Department at Sandia National Laboratories in Albuquerque, NM. His work is primarily on the Chain of Custody project, in which he is the Hardware and Operations Design Lead. He is primarily concerned with the development and evaluation of candidate technologies to be used as monitoring nodes at the Chain of Custody Test Bed. Last year, Jay participated in the Next Generation Working Group on U.S.-China Nuclear Relations, a function of the Center for Strategic and International Studies (CSIS) Project on Nuclear Issues (PONI). Jay graduated with a B.S. in Computer Engineering from Rose-Hulman Institute of Technology and an M.S. in Electrical and Computer Engineering from Carnegie Mellon University, where he wrote a Master’s thesis on damping of mechanical resonators fabricated in a CMOS-MEMS process.
Justin Fernandez is a Senior Member of the Technical Staff at Sandia National Laboratories. Justin’s experience and expertise lies at the intersection of technology and policy, with a focus on international nuclear relations and arms control. For the past two years he has led test and evaluation activities between three national laboratories for nuclear monitoring and transparency technologies geared towards supporting future arms control initiatives. Prior to his current position, Justin worked for three years on testing and evaluating the compatibility of Sandia developed technologies with US Air Force and NATO aircraft platforms. Justin obtained his B.S. and M.S. in Mechanical Engineering from Rutgers University and Georgia Institute of Technology respectively.
Dr. Sharon DeLand is a System Analyst in the Nuclear Monitoring and Transparency Department at Sandia National Laboratories. She received her doctorate in experimental condensed matter physics from the University of Illinois in 1991. Sharon’s current research interests include developing and evaluating technical approaches for monitoring arms control agreements, especially approaches focused on item accountability. Her work focuses on systems approaches that integrate technical monitoring objectives with policy perspectives and operational constraints. She also applies systems analysis to the modeling and simulation of international relations, with an emphasis on nonproliferation and arms control.
The opinions expressed in this paper are the authors’ own and do not reflect the opinions or official policy of Sandia National Laboratories, the National Nuclear Security Administration, or the United States Government.
- New Strategic Arms Reduction Treaty (New START), U.S. Department of State, Signed April 8, 2010, https://www.state.gov/documents/organization/140035.pdf. ↩
- Remarks by President Barack Obama, Prague, Czech Republic, April 5, 2009, http://www.whitehouse.gov/the_press_office/Remarks-By-President-Barack-Obama-In-Prague-As-Delivered/. ↩
- Nuclear Posture Review Report, U. S. Department of Defense, April 2010, http://www.defense.gov/npr/docs/2010%20Nuclear%20Posture%20Review%20Report.pdf. ↩
- New Strategic Arms Reduction Treaty (New START), U.S. Department of State, Signed April 8, 2010, https://www.state.gov/documents/organization/140035.pdf. ↩
- The production stage is theoretical only, since the United States no longer produces nuclear weapons, and is provided here for completeness of the abstraction. ↩