Public Safety Drone ROI: Three Legs of the Business Case

Return on investment in public safety is a strange kind of math.

In most domains, ROI is a clean calculation: dollars in, dollars back, payback period, done. In emergency response, the return includes things no finance director wants to price. Lives saved. Officers not injured. A missing child found before hypothermia got to her. These are the outcomes that justify the program to a community, and they’re also the outcomes that make a straight cost-benefit analysis feel reductive, because reducing a rescued grandmother to a line item is, rightly, something we don’t do.

But the math still gets done. It gets done by council members evaluating a budget request. It gets done by commissioners deciding whether to renew a program. It gets done by city managers trying to justify next year’s capital plan. And the agencies that walk into those rooms with defensible numbers get funded. The ones who walk in with stories alone, too often, do not.

This is a breakdown of the numbers that actually survive that scrutiny: where they come from, what they mean, and how to present them without overselling or underselling the case.

The three legs of the ROI stool

A credible public safety UAS business case rests on three legs. They each answer a different question, and the strongest cases address all three.

Leg one: direct operational cost reduction. What spending is displaced, reduced, or avoided by adding drones to the tool kit?

Leg two: outcome improvements. What gets measurably better, response times, time-to-locate, case clearance, search success?

Leg three: risk and exposure reduction. What liability, officer-injury, or civilian-harm exposure is reduced, and what’s that worth?

Each leg has defensible numbers behind it. We’ll walk through them in that order. For the internal-champion framing of how these numbers turn into an approved budget line, pair this with our UAS platform business case playbook.

Leg one: operational cost reduction

The cleanest savings line is helicopter displacement.

Acquisition cost for a police or SAR helicopter runs from $500,000 on the low end to $3 million and up. ACLU analysis cites this range, and individual agency purchases confirm it. Tulsa PD bought a new Eurocopter AS350B2 in 2013 for roughly $2.5 million including law enforcement upgrades. A new Bell 206 JetRanger configured for SAR lists around $700,000. These are before outfitting, hangar, insurance, and the pilot roster.

Operating cost is where it gets interesting. Public figures range from $200 to $400/hr (ACLU baseline) to $600/hr (Long Beach PD’s estimate for its AS350B2, exclusive of pilot salary) to $800/hr (MeriTalk cited figure for fully-loaded operations). Long Beach publicly estimated spending nearly $414,000 annually operating its helicopter about 650 hours per year at the $600/hr figure, and that was conservative, excluding pilot pay, training, fuel, and irregular maintenance.

Drone operating cost, at the same level of transparency, lives in a different universe. Mesa County (Colorado) Sheriff’s Office, an early adopter, publicly estimated their per-hour drone operating cost at approximately $25, almost entirely driven by propeller and battery replacement. Four hundred flight hours a year costs them around $10,000. That’s not a typo. It’s a difference of roughly two orders of magnitude on the hourly rate.

No serious program claims drones replace helicopters outright. They don’t. Manned aviation still wins on endurance, payload, pursuit speed, and large-area coverage. The real math is about displacement: of the hours an agency currently flies manned aircraft, what percentage could be handled by a drone at $25/hr instead of $600? For most small-to-mid-sized agencies, the answer is a significant portion of search support, scene documentation, perimeter coverage, and situational awareness work.

Polk County (Florida) Sheriff’s Office publicly reported saving “tens of thousands of dollars” over eighteen months by using drones in place of their manned aviation assets across 750 missions. That’s a figure the Sheriff was willing to put his name to on the record, in a context where over-claiming would be noticed.

A 2026 note on how these costs increasingly come to market. The public safety UAS industry has shifted meaningfully toward bundled service models (software, aircraft, maintenance, and scheduled replacements included in an annual subscription) rather than one-time hardware purchases. For many buyers, this is a better fit financially than a capital purchase: it converts a lumpy every-four-years hardware budget into a predictable operating expense, dissolves the replacement-cycle headache, and tends to move through city councils more easily than capital requests. The tradeoff is flexibility (you’re tied to one vendor’s roadmap), so the right question isn’t which model wins universally, but which better matches your procurement posture and appetite for year-three uncertainty.

The second operational savings line is personnel hour displacement.

Chula Vista PD’s DFR program, the original and still among the most transparent about its data, has cited more than 4,000 instances of drones handling a call and clearing it without ground-unit dispatch. Across multiple programs, data from 2025 shows DFR reducing unnecessary patrol dispatches by roughly 15% to 20%. Miami Beach reported clearing 41% of DFR-responded calls without officers over a 22-week window. These are not marginal numbers. At a fully-loaded officer cost of $85 to $120 per hour (varies by agency and benefits load), a 15% reduction in unnecessary dispatches across a mid-sized department represents real money, and the officer hours it frees up are hours that become available for calls where a human response is actually needed.

One underrated cost line: accident scene reconstruction. The traditional timing is roughly four hours of officer time to photograph and measure a collision scene, with the road closed for the duration. A drone running photogrammetry or LiDAR mapping does the same work in about fifteen minutes, typically at half to three-quarters less cost than contracted total-station survey work, and modern videogrammetry tooling can render the scene models in minutes rather than days. For an agency handling any meaningful volume of crashes, that’s hours of officer time, road-closure exposure, and secondary-collision risk recovered per incident.

Leg two: outcome improvements

This is the leg that matters most to the community and is the hardest to put a single dollar figure against. The numbers still exist, they just have to be translated into outcomes rather than savings.

Before the specific metrics, a scale marker. The Las Vegas Metropolitan Police Department flew more than 10,000 drone missions in 2025, the highest annual volume of any public agency in the United States, and has publicly projected roughly 20,000 missions in 2026 at its current cadence of about 1,700 flights per month. The department operates 75 aircraft across 13 rooftop skyports, with the Skydio X10 as the flagship docked aircraft. Whatever else can be said about the program (and there are legitimate transparency questions about its hybrid public/private funding model), it puts to rest any remaining doubt that mid-to-large DFR operations are still experimental. They are operational infrastructure, running at a scale that rivals regional manned-aviation units. For the broader shift this represents, see our piece on how drones are transforming public safety.

Response time. DFR programs have normalized response speeds that were impossible before. Skydio’s integrated DFR Command system reports aircraft launching in under 20 seconds and reaching incident scenes in under 90 seconds. Flock’s DFR platform publicly cites an 86-second average response time, with 78% of calls resulting in a drone on-scene before ground units arrive. Lakewood PD, Colorado, reported drones arriving before ground units 80% of the time in its first six months of operation, contributing to 131 arrests. The Redmond PD, Washington, runs DFR with sworn officers as remote pilots and can reach anywhere in its 14-square-mile service area in 90 seconds or less.

These are not vendor brochure numbers. They are numbers published by operating agencies with public dashboards.

Time-to-locate. In search and rescue, this is the variable that most directly determines outcomes. The NASAR-reported window is stark: the probability of finding someone alive drops sharply after 72 hours, and for many subject profiles (hypothermia-exposed, medical conditions, pediatric) the real window is much tighter than that. Specific cases give the shape: Cass County, Missouri, located a missing 93-year-old woman in four minutes after launching a drone with thermal imaging, after ground searchers had been unsuccessful for over three hours. A UK rescue located a missing 13-year-old who had taken refuge hidden under vegetation; the attending officer later stated she would not have been found before morning without the drone. These are single cases, but they follow a pattern: when thermal-capable drones are available, the window from “ground search exhausted” to “subject located” collapses from hours to minutes in a meaningful percentage of cases.

Call clearance and force multiplication. The 15 to 20% unnecessary-dispatch reduction figure above is also an outcome metric, not just a cost metric. Those freed ground units are now available for priority calls they would otherwise have been diverted from. In cities running severe staffing shortages (San Francisco’s department ran hundreds of officers short when it launched DFR and RTIC), this capacity multiplication is the program’s most important outcome, more than any cost line.

Documented rescues. DJI’s “Drones for Good” tracker, which logs drone-assisted rescues documented in news reporting or by agencies, passed 500 rescues in October 2020 and has grown substantially since. That tracker is almost certainly a significant undercount, most rescues never get the press coverage that makes it into a manufacturer database. But as a floor for the documented outcome volume, it establishes that we are past the anecdote phase and firmly into the pattern-of-evidence phase.

Leg three: risk and exposure reduction

The third leg is where the business case most often goes underdeveloped, because it deals in counterfactuals. We don’t know, precisely, how many officer injuries didn’t happen because a drone cleared the building first. We know the number is greater than zero, and directionally significant, and that’s enough to include in the case, but it has to be framed honestly.

The categories that belong here:

Officer exposure reduction. Building clears, structure searches, perimeter sweeps, open-field tracking of armed subjects, all of these traditionally put officers into environments where the threat is real and the visibility is poor. Drones flying ahead of entry, or substituting for an entry entirely, change the risk profile of the operation. Dallas PD’s “Drone Clear” program has demonstrated suspects surrendering specifically because a drone flew into their position; Atlanta has on the record an apartment-entry situation that ended peacefully after a drone confronted the subject rather than an officer.

De-escalation outcomes. Chula Vista PD’s chief has described DFR as one of the department’s most effective de-escalation tools, specifically because an aerial view frequently confirms that a “weapon” report was inaccurate or that a scene is less kinetic than initially described. A call that would have drawn a high-intensity response becomes a lower-intensity one. The value of preventing a bad outcome is, by nature, hard to quantify; that doesn’t make it zero.

Liability profile changes. Agencies running defensible documentation (time-stamped flight logs, role-based access, court-admissible video) have measurable advantages in post-incident review, FOIA response, and civil litigation. This isn’t flashy, but ask any city attorney what a drone mission record that was properly generated and retained is worth in the event of a use-of-force case. The answer is: it depends, but it’s never zero, and sometimes it’s decisive.

Officer wellness and retention. The less-discussed leg. Agencies consistently report that officers, particularly younger officers, prefer working in departments that provide modern technology to do their jobs. In a recruiting environment where most departments are short, the retention and recruiting effect of a credible technology program is real. Not easily quantified. Not zero.

The grant funding offset

One thing every credible business case should acknowledge: the sticker price is not the out-of-pocket cost for most public safety agencies. A meaningful fraction of drone program costs can be offset by grants, and the landscape is well-developed enough in 2026 that this is a routine consideration rather than a long shot.

The primary federal pathways: FEMA’s Assistance to Firefighters Grants (AFG) for fire departments; DHS’s Homeland Security Grant Program (HSGP), which includes the State Homeland Security Program (SHSP) and Urban Area Security Initiative (UASI); DHS Operation Stonegarden (OPSG) for border-adjacent agencies; DOJ COPS grants for law enforcement; and, for specific use cases, FEMA Port Security and state Section 402 Highway Safety program funds. Private sources include the EQT Foundation (volunteer SAR and firefighting), the Firehouse Subs Public Safety Foundation, and regional programs like the Hero Fund. For DFR-specific equipment, the DHS Authorized Equipment List covers these systems under AEL 03OE-07-SUAS, which makes HSGP eligibility explicit.

In practice: a credible year-one capital budget should be presented two ways, gross cost and out-of-pocket cost assuming realistic grant participation. The delta is often 30 to 60%. Budget holders who don’t see both numbers are being underserved by whoever prepared the case. Our government drone procurement guide walks through the grant-and-procurement side in more depth.

How to build a business case that survives

A few things separate the business cases that get funded from the ones that don’t.

Present three-year and five-year projections, not just year one. Year one is the expensive year. The operational savings curve, the grant-funded offsets, the outcome metrics, all of them compound past year one. A council member looking only at year-one numbers is looking at the worst version of the case. Model it at three and five years, and show the turn.

Separate cost-avoidance claims from capability claims. A weak case says “this saves you money.” A strong case says: here’s the capability this gives you that you don’t currently have; here’s its cost; here’s how it compares to the alternatives; here’s what the alternatives cost and why they aren’t a real option for us. Budget holders respect honesty about limits. They distrust vendors, and internal champions, who present only upside.

Anchor the outcome case in your own jurisdiction. “DFR reduces unnecessary patrol dispatches by 15 to 20%” is a national figure. “Last year our agency ran X calls that a DFR program would likely have cleared without officer dispatch, at a fully-loaded cost of Y hours” is your figure. It’s the one that wins the room. Build it from your own CAD data before you walk in.

Don’t oversell the lives-saved line. This is the most important piece of advice in this entire article. You do not need to claim the program will save a specific number of lives. You need to show that the capability meaningfully affects time-to-locate, scene awareness, and officer exposure, and let the community evaluate what that’s worth. Overclaiming here gets agencies in trouble when the program’s first year turns out to have zero headline rescues. Underclaim, and let the outcomes speak.

The honest close

ROI in emergency response will never be a clean finance-office calculation, and pretending otherwise does the program a disservice. The right frame is not “this pays for itself.” It’s “this gives the agency a capability at a cost that’s defensible against every alternative, with outcome evidence drawn from comparable agencies and grant-funded offsets that reduce year-one exposure substantially.”

The agencies running credible programs have done that math and walked into the room with it. The numbers in this article, helicopter hour displacement at roughly 20-to-1 on operating cost, dispatch reduction in the 15 to 20% range, time-to-locate measured in minutes rather than hours, grant-funded offsets of 30 to 60% for well-positioned applicants, are defensible, public, and sourced to agencies willing to put their name behind them.

The work, now, is translating them into the version of the case your own jurisdiction needs to see.