August 10, 2015

The $100 Billion Question

The Cost Case for Naval Uninhabited Combat Aircraft

Executive Summary

  • The Navy is already beginning to examine options to replace the F/A-18E/F Super Hornet, which will retire in the mid-2030s.
  • The replacement of the F/A-18E/F with a future naval aircraft (FNA) represents a major opportunity to shape the Navy’s future carrier air wing to respond to emerging challenges but must take into account budget constraints.
  • This analysis examines the potential cost differences in a notional human-inhabited (“manned”) FNA compared to an uninhabited (“unmanned”) FNA using three cases (conservative, moderate, and aggressive).
  • All cases are structured to support the Navy’s current Optimized Fleet Response Plan, which calls for two deployed carrier strike groups at all times, with the ability to temporarily surge up to six carriers.
  • All uninhabited aircraft cases generate major cost savings, achieved by avoiding the costs of recurring pilot training.
  • Because the uninhabited aircraft software flies the aircraft, costly flying hours are not needed to train pilots to control the aircraft and maintain their skills. The aircraft flies itself, with the remote pilot (“operator”) in a decision-maker role, providing mission-level command.
  • With fewer hours flying, fewer aircraft are needed.

Range of Potential Savings

We assess the minimum cost avoidance at about $30 billion over 30 years, with a more likely cost avoidance in excess of $100 billion.

  • Because the cost of FNA is not yet known, we use a parametric cost model to examine a range of possible costs, depending on assumptions regarding cost of a next-generation aircraft.
  • The Navy can achieve this cost avoidance only by replacing manned aircraft with uninhabited aircraft.
  • These substantial savings could then be re-invested into other Navy priorities, such as additional ships, submarines, or other aircraft programs.
  • These savings do not include real and highly substantial funds that could be saved by reductions in the initial pilot training infrastructure and supporting depot infrastructure.
  • This is a “think piece” with many simplifying assumptions and is intended to illustrate the principles of what is different with uninhabited aircraft.


The Navy is already beginning to examine options to replace the F/A-18E/F Super Hornet, which will retire in the mid-2030s.

The replacement of the F/A-18E/F with a future naval aircraft (FNA) represents a major opportunity to shape the Navy’s future carrier air wing to respond to emerging challenges but must take into account budget constraints.

This analysis examines the potential cost differences between a notional human-inhabited FNA and an uninhabited FNA.

  • This analysis isolates cost issues alone. It assumes that technology has matured to the point where an uninhabited aircraft with a high degree of autonomy and communications links to human operators would be at least equal in capability to a human-inhabited aircraft.
  • Uninhabited aircraft also have significant operational advantages over human-inhabited aircraft due to their ability to operate beyond the endurance limits of human pilots, which this analysis does not include.

The U.S. carrier fleet is trained and organized for two readiness objectives: To generate continuously a certain number of deployed carrier groups – currently two carrier strike groups, and to generate temporarily a larger surge deployment capability – up to six carrier strike groups.

Each alternative force examined must yield sufficient aircraft, including training, depot, and attrition aircraft, to meet both the steady-state (two carriers) and temporary surge (six carriers) requirements.

The Navy has 11 aircraft carriers, supporting 10 carrier strike groups.

The carriers rotate through a regular 36-month readiness cycle, moving through periods of maintenance, pre-deployment preparation, deployment, and then a lengthy sustainment period where they are available for surge, as needed.

At 25 years, a carrier needs to enter deep maintenance for a refueling and complex overhaul of its nuclear reactor, a process which takes approximately four years.

Under the Navy’s new Optimized Fleet Response Plan, at any given point in time there are roughly two carriers deployed, four available for surge, two in maintenance, two preparing for another deployment, and one in deep maintenance.

What’s Different for Uninhabited Aircraft?

First-generation uninhabited air vehicles (UAVs) were remotely controlled by a pilot on the ground, but newer UAVs are highly automated. Remote pilots (“operators”) command the aircraft and tell it where to go, and the aircraft flies itself.

Human-Inhabited Aircraft

Onboard pilots directly control the plane. Each pilot must be trained individually, and pilot skills degrade over time. Large amounts of flying hours are needed to retain currency. Training simulators are useful, but there is no substitute for hands-on, in-the-cockpit experience.

Highly Automated UAVs

Remote pilots command the UAV at the mission level. Using onboard software, the UAV flies itself. All UAVs are equally skilled and capable the moment they are completed, and their skills don’t improve or degrade over time. No additional flying hours are needed.

We assume that cost differences unique to human-inhabited vs. uninhabited aircraft variants are sufficiently small that they can be excluded. Differences in life-cycle cost will therefore be driven by the number of aircraft purchased and the way these aircraft are operated.

Aircraft Life-cycle Costs

Total costs throughout an aircraft program consist of research and development (R&D), investment (aircraft production), operating and support (O&S), and disposal. Since R&D costs are driven primarily by mission requirements, which would be the same for both aircraft variants, this study will focus on investment costs and operating and support costs. 

Life-Cycle Cost Differences

Cost CalculatorDifference between Inhabited and Uninhabited AircraftTreatment in Study
Research and Development (R&D)Marginal DifferenceExcluded from Study
InvestmentMay vary significantly depending on the number of aircraft purchaseCentral to study
Operating and Support (O&S)May vary significantly depending on the number of hours flowCentral to study
DisposalSame cost per plane; small percentage of total cosExcluded from Study

Human-Inhabited FNA

First, we estimate the cost of a notional human-inhabited FNA as a base case. In order to meet the requirements of two carriers deployed steady-state and the ability to surge up to six, 463 total aircraft are needed.

The number of aircraft is determined by:

  • 24 aircraft per air wing for all 10 carrier air wings (carrier in deep maintenance does not require an air wing);
  • 144 aircraft to meet six wing surge demand continuously present on carriers;
  • Training squadron = 37% of carrier aircraft = 89 aircraft;
  • Aircraft at depot maintenance = 15% x (carrier + training aircraft) = 49 aircraft;
  • Attrition occurs at an average rate of three losses per 100K hours flown = 85 aircraft over a 25-year fleet service life; and
  • 31 flying hours per aircraft per month for aircraft in the 10 air wings; 23 flying hours per aircraft per month in the training squadron.

Human-Inhabited FNA Costs (modest cost increase)

Excluding R&D and disposal costs, total program life-cycle costs will consist of the costs to procure the aircraft plus the O&S costs to operate and support them over a 25-year lifespan.

Since the precise costs of an FNA are not known, this analysis uses a parametric cost model, based on actual F/A-18E operating costs and anticipated F-35C costs. Three cases are presented, representing a range of estimates for the anticipated increase in costs for a next-generation FNA relative to the F-35C:

  1. No increase in costs relative to the F-35C;
  2. A modest increase in costs relative to the F-35C; or
  3. A significant increase in costs relative to the F-35C, equivalent to the increase in costs from the F/A-18E to F-35C.

Since the key variable is the difference between an uninhabited and a human-inhabited FNA, a rough order-of-magnitude estimate of FNA costs is sufficient to understand the potential for cost savings.

Range of Costs for Human-Inhabited FNA

F/A-18E CostFNA Costs (No Cost Increase)

FNA Costs (Modest Cost Increase)

FNA Costs (Significant Cost Increase)
Actual F/A-18E costsIf FNA costs equaled F-35C costs

If the cost increase from the F-35C to FNA were half as much as the cost increase from the F/A-18E to F-35

If the cost increase from the F-35C to FNA were as much as the cost increase from the F/A-18E to F-35C
Number of Aircraft Procured463463


Unit Procurement Cost$75 M$130 M

$175 M

$220 M
Total Procurement Cost$34.7 B$60.2 B

$81.0 B

$101.9 B
Cost per Flying Hour$15K/hr$29K/hr


Total Operating and Support Cost$45.5 B$84.9 B

$118.3 B

$151.7 B
Total Program Life-Cycle Cost*$80.2 B$141.5 B

$199.3 B

$253.6 B

Uninhabited FNA – Conservative Case

The conservative case estimates the cost savings from a single change: a reduction in the number of aircraft required in the training squadron. Because the UAV flies itself, remote pilots/operators do not need to train on aircraft handling skills. Moreover, because there is no “seat-of-the-pants” feel gained from flying the aircraft, a significant fraction of mission training can be done via simulator, reserving actual aircraft flying for final exercises.

In the conservative case:

  • Training aircraft are reduced to 16 aircraft. These are used for handler and maintainer training (who require hands-on experience with the aircraft) and squadron-level joint training exercises with manned aircraft.
  • This results in a reduction in the number of aircraft at the depot.
  • A reduced training squadron means fewer hours flown and a corresponding reduction in the number of attrition aircraft.
  • The number of aircraft allocated to each carrier air wing remains unchanged, as does the number of flying hours per month.
  • 144 aircraft meets the six wing surge demand continuously present on carriers.

The full report is available online.

Download PDF

  1. With the retirement of the USS Enterprise (CVN-65) in 2012, the Navy currently has only 10 carriers in the fleet. Provided the Navy overhauls the USS George Washington (CVN-73) at its scheduled 25-year refueling and complex overhaul, the number of carriers will return to 11 when the USS Gerald R. Ford (CVN-78) enters service in 2016. The Navy has a statutory requirement under Section 1011(a) of the FY2007 National Defense Authorization Act to maintain no less than 11 operational carriers. The Navy is currently operating under a temporary waiver from Congress, authorized in the FY2010 National Defense Authorization Act, to operate 10 carriers until CVN-78 is commissioned. For additional background, see Ronald O’Rourke, Navy Ford (CVN-78) Class Aircraft Carrier Program: Background and Issues for Congress, Congressional Research Service, Washington, DC, June 12, 2015.
  2. Admiral Bill Gortney, USN, “Navy Optimized Fleet Response Plan” (26th Annual Surface Navy Association National Symposium, Crystal City, VA, January 2015).
  3. As of 2015, the Navy program for F/A-18E/F is 563 aircraft. This actual program number differs from the “steady state ideal” number (463) for potentially a number of reasons. The most likely reason is that the F-35C is perhaps not being produced as quickly or at as great a rate as would be ideal for the air wing, especially given the additional flying hours on the F/A-18E/Fs from supporting operations in Iraq and Afghanistan. This means that greater numbers of E/F models are needed to compensate for shortfalls. The steady state number also does not account for the (relatively minor) effects of aircraft being introduced into the air wing over time, as opposed to an instantaneous total quantity delivery.
  4. Derived from Navy Visibility and Management of Operating and Support Costs (VAMOSC) data. Since the actual number of aircraft in the training squadron (“fleet replacement squadron“) varies over time, 37% is a representative mid-range value.
  5. Actual number of aircraft in depot maintenance varies over time. 15% is a representative figure. Since the same depot percentage is used for all cases, however, a slightly higher or lower depot percentage would not significantly change the difference in costs between cases.
  6. Based on F/A-18 Class A mishap rate, 1990 –2013. Edward Hobbs, “Comparison of Aviation Mishap Rates for Hornet Squadrons During Periods of Extended Reduced Flight Hours With Periods of Normal Flight Operations,” Navy Safety Center, 9,
  7. Representative monthly flying hours, derived from VAMOSC data.
  8. More precisely, we calculate the cost to procure and operate two squadrons per each of two deployed carriers for 25 years ( = 100 “deployed squadron years”) – along with all the other aircraft needed to sustain this deployed capability and a six air wing surge capability. This will take longer than 25 years since production takes more than a decade to complete; hence there will be a variable number of squadrons extant at any given time.
  9. Costs calculated using the Navy’s Operating and Support Cost Analysis Model (OSCAM).
  10. This is how training is already done for highly automated UAVs like the RQ-4 Global Hawk, where 100 percent of flight control training and approximately 40 percent of mission training is performed via simulator, significantly reducing training flying hours. See Paul Scharre, “Can Automation Reduce Training Costs: A Preliminary Assessment Based on a Comparison Between U.S. Air Force Manned and Unmanned Aircraft Pilot Initial Qualification Training,” Center for a New American Security, Washington, DC, October 13, 2014,


  • Daniel Burg

  • Paul Scharre

    Executive Vice President and Director of Studies

    Paul Scharre is the Executive Vice President and Director of Studies at CNAS. He is the award-winning author of Four Battlegrounds: Power in the Age of Artificial Intelligence...

  • Podcast
    • April 29, 2021
    Fire and Ice

    In this week’s edition of the SpyTalk podcast, Jeff Stein goes deep on the CIA’s looming eviction from Afghanistan with Lisa Curtis, a longtime former CIA, State Department an...

    By Lisa Curtis, Jeff Stein, Jeanne Meserve & Alma Katsu

  • Commentary
    • Texas National Security Review
    • June 2, 2020
    The Militarization of Artificial Intelligence

    Militaries are racing to adopt artificial intelligence (AI) with the aim of gaining military advantage over competitors. And yet, there is little understanding of AI’s long-te...

    By Paul Scharre

    • Commentary
    • Wired
    • May 19, 2020
    Are AI-Powered Killer Robots Inevitable?

    In war, speed kills. The soldier who is a split second quicker on the draw may walk away from a firefight unscathed; the ship that sinks an enemy vessel first may spare i...

    By Paul Scharre

  • Commentary
    • The New York Times
    • February 12, 2020
    The Iranian Missile Strike Did Far More Damage Than Trump Admits

    Over 100 American soldiers have been treated for traumatic brain injuries following Iran’s missile strike on Al Asad Air Base in western Iraq. The strike came in retaliation f...

    By Loren DeJonge Schulman & Paul Scharre

View All Reports View All Articles & Multimedia