Every one of these cities cut spending the same way: buy better information about what is actually happening, so the same crews and assets do more useful work. That is the illegal-dumping problem too.
The question a city council actually asks is not "does this technology work." It is "does this cost less than the problem." And the honest answer for most municipal technology is that it depends entirely on whether the technology changed a decision.
A new system that makes the same work slightly faster saves nothing, because the saved time gets absorbed. What saves money is a system that stops a crew from driving somewhere for nothing, stops a truck from being purchased to replace a truck that was idle, stops an asset from running when nobody is there to receive it, or stops a manager from approving overtime that nobody needed to be approved.
Six examples follow, each one drawn from a primary source: a city program page, a city audit, a university report, or a city official on the record. Every number below has a link to where it came from. We have left out programs whose savings claims could not be traced to a primary document, including several smart-waste sensor vendor case studies that report savings with no independently published city figure behind them.
The short list
| City | What changed | Documented result |
|---|---|---|
| Chicago | Risk-based inspection scheduling for restaurants | Critical violations found 7.5 days earlier |
| Los Angeles | Citywide LED streetlight conversion | ~$10M/yr 64% energy reduction |
| Pittsburgh | Adaptive traffic signals, no new road capacity | 40% less wait time ~26% less travel time |
| Philadelphia | Automated meter reading, 470,000+ meters | ~60 positions plus fleet and overhead savings |
| Franklin, TN | Waste route optimization | $1,083,812/yr 37% of operating cost |
| Baltimore | Digital time entry plus overtime oversight | $66M exposed 18-month audit window |
Six different functions: restaurant inspection, street lighting, traffic, water metering, waste collection, and police payroll. Six different technologies: predictive scheduling, LED hardware, real-time control, radio telemetry, route optimization software, and a modern HR system. The savings mechanism is the same in all six.
1. Chicago: same inspectors, better order
Under the same inspection staffing, establishments with critical violations were found an average of 7.5 days earlier.
The city inspects more than 15,000 food establishments with roughly three dozen inspectors, which works out to about 470 establishments per inspector. The inspection schedule was set by how long it took to work through the list, so every establishment got visited on the same cadence whether it was clean or not.
The city ran a double-blind study: inspectors visited restaurants they had no prior reason to suspect, and researchers compared where violations turned up against where the existing schedule would have sent them. Under normal operations, inspectors found about 69% of critical violations in the first half of their shift. Reordered by risk, that number rose to roughly 80%, and establishments with critical violations were found 7.5 days earlier on average.
A city does not need more cleanup crews. It needs the crews it has to work on the piles most likely to be real, most likely to be growing, and most likely to be there again next week. This is the same decision Chicago made, made with aerial imagery instead of a predictive model. Every pile that gets found and worked before it grows is a trip that happens once instead of twice, and a pile that nobody reported is a trip that would not have happened at all.
2. Los Angeles: the savings come after the invoice
Converting the streetlight fleet to LED cut lighting energy use by 64%, saving roughly $10 million a year.
Los Angeles converted more than 140,000 streetlight fixtures to LED, finishing the second phase of the program between 2014 and 2019 and bringing the city to about 98% LED coverage. The measured result is 114 gigawatt-hours of annual energy savings and roughly 67,000 metric tons of avoided carbon emissions.
The structure of this one is worth naming, because it is the structure a city procurement office will ask about. There was a capital cost up front, and then a permanent reduction in the operating line every year afterward. The savings are not a one-time event. They are an annuity.
Aerial detection is the capital-and-annuity shape too. A city pays for coverage, and then every verified work order that turns out to be real, and every missed pile that gets found before it becomes a three-week problem, is operating savings that keep accruing. A contract that only produces a monthly report does not have this property. A contract that makes the existing crews measurably more productive does.
3. Pittsburgh: better signals, no new lanes
Adaptive traffic signals cut vehicle wait time 40% and travel time nearly 26%, with no new road capacity.
Carnegie Mellon built a system where traffic signals communicate with each other and adjust to actual conditions in real time, then piloted it in Pittsburgh's East Liberty neighborhood with the city and East Liberty Development Inc. The pilot cut vehicle wait time by 40%, travel time by nearly 26%, and projected vehicle emissions by 21%.
What matters for a budget conversation is what the city did not have to do. The Pittsburgh mayor described the result as making the traffic system work more efficiently "without having to resort to expensive widening roads, eliminating street parking, or re-routing." The capacity was already there. It was being wasted by signals that did not know what was happening three blocks away.
Illegal-dumping capacity already exists. The trucks are bought, the crews are hired, the overtime is budgeted. What does not exist is reliable information about where the piles are, so a good share of that capacity drives to empty lots and comes back. Adding coverage to a city that already owns the trucks and the crews is the traffic-signal move: use the capacity you have, better, without asking for more.
4. Philadelphia: the meter reports itself
Automated meter reading eliminated about 60 meter reader positions, plus fleet, overhead, and workers' compensation costs, with no layoffs.
Philadelphia's Water Department deployed automated meter reading across more than 400,000 meters, with three mobile units reading over 400,000 meters a month at a read success rate above 99% and a read accuracy rate of 99.7% or better. The department later completed a full Advanced Metering Infrastructure upgrade across roughly 470,000 meters, giving every household in the city daily usage data over a secure radio signal and leak alerts before a bill arrives.
The savings came from the labor and fleet costs that vanished with the physical act of walking or driving to each meter: about 60 meter reader positions eliminated, along with associated vehicle expenses, overhead, and workers' compensation costs. The department made the point that it did this without layoffs, reassigning displaced staff or letting positions go through attrition, and used the freed capacity to cut call waiting and abandonment rates instead of shrinking the service.
This is the closest structural match. In both cases, the city did not cut the service. It removed the manual round trip that produced the information, because a sensor could produce it better. A crew that has to drive to a location to find out whether a pile exists is doing a meter reader's job. Aerial imagery does that pass from the air, and the measurement the city gets back is the read.
5. Franklin, Tennessee: fewer miles, fewer trucks, less labor
Route optimization cut the city's waste collection operating costs by 37%, or $1,083,812 a year.
Franklin was planning collection for about 38,000 households using paper maps, GPS units, spreadsheets, pens, and driver knowledge. After deploying route optimization software and a driver navigation tool, the city consolidated residential collection from a five-day to a four-day work week, cut the street sweeping cycle from seven weeks to three, and reduced its annual waste operating costs by 37%, which the city puts at $1,083,812, attributed directly to fewer miles driven, less truck usage, and lower labor cost.
The city's routing and operations analyst put the emphasis on repeatability rather than a single optimization run. The value was in being able to re-optimize as the city grew, and in showing dispatchers and drivers the same plan.
| Franklin, TN | Before | After |
|---|---|---|
| Annual waste operating cost | Baseline | $1,083,812 lower |
| Residential collection schedule | 5 days | 4 days |
| Street sweep cycle | 7 weeks | 3 weeks |
| Households served | ~38,000 | ~38,000 |
This is the waste-collection example, and it is the one that maps most directly. Franklin's problem was that its routes were built on what was believed about pile locations rather than what was actually there. The same is true of a reactive 311 dispatch queue. A work order that points to the wrong spot, or to a spot that is now empty, sends a truck on a trip that produces nothing and returns nothing. Know where the piles are, batch by area and by equipment, and the route gets shorter and the truck utilization goes up. The savings line in Franklin's case was 37% of operating cost. The mechanism was better location data.
6. Baltimore: the audit that made the spending visible
An audit of police overtime found $66 million spent over 18 months without adequate supervisory review, and the fix was digital time entry plus real oversight.
Baltimore's Office of Legislative Audits reviewed the Baltimore Police Department's overtime policies, procedures, and activity for December 2020 through June 2022 and found $66 million in overtime costs that had been paid without adequate supervisor oversight, a significant share of it from individuals who had received little or no supervisory review. The audit identified the enabling conditions: the legacy payroll system had weaknesses that allowed time entry and leave tracking to be abused, and supervisory enforcement of overtime policy was inconsistent.
The department's own response is the useful part. The audit period coincided with the rollout of Workday as the city's new HR and payroll system, configured with internal controls that moved the department out of paper-based time entry into digital entry with employee self-service tools. The audit found no weaknesses in the Workday configuration or in the policy changes around time tracking. The department also moved to a third-party off-duty management system, launched an accessible overtime dashboard, began regular internal audits of overtime activity, and started holding quarterly reviews of General Fund budgets including overtime.
Here the mechanism is visibility, and it is worth separating from the other five because it is the one that gets skipped. Baltimore did not save $66 million with better technology. It spent $66 million without noticing. The technology is what made the spending visible enough to manage, and the management change is what turned visibility into a control. For a city operating on a reactive 311 queue, the same thing applies: the cost of dry runs and wrong-location work orders is real money, but it is invisible until someone measures it. A city cannot cut a line item it cannot see, and it cannot defend a contract that has no measurement behind it.
The pattern across all six
Strip away the specifics and every one of these programs does the same three things.
- It replaces a guess with a measurement. Chicago replaced a calendar with a risk score. Franklin replaced driver memory with optimized routes. Philadelphia replaced a physical meter read with a transmitted read. Baltimore replaced an unverifiable paper timesheet with a digital entry that cannot be quietly altered. In each case, the city already had the assets and the people. It just did not know what it was doing with them.
- It changes a decision, not just a speed. Faster is not the goal. The goal is that a truck does not dispatch to an empty lot, that an inspector walks into the restaurant most likely to have a problem, that a signal waits when the cross street is empty, that an overtime request gets looked at by somebody whose job it is to look at it. Every one of these programs moved money because it changed what somebody decided to do.
- It saves operating cost, not just time. The savings are in trucks, miles, labor hours, overtime, and energy. Those are the lines a city actually cuts. A program that only produces a better report saves nothing.
Why illegal dumping is the same shape
Illegal dumping has all three properties, and the numbers we have measured in Oakland make them concrete.
It replaces a guess with a measurement. A 311 request is a report, not a survey. It records where someone thought the pile was and roughly when they noticed it. It does not record whether the pile is still there, what is actually in it, or whether anything has been dumped on top of it since. In our verification of 1,036 open City work orders against 1,380 accepted piles from 19 coverage flights, 96% of work orders carried some kind of exception a crew would only discover by going: 37% had no pile within 100 yards at all, 31% pointed at the wrong location, and 27% named only part of what was on the ground.
It changes a decision, not just a speed. That 37% is a dispatch decision, and it is the most expensive line in the set. A crew sent to a lot with nothing on it produces no cleanup, and the next request for that same spot has to start over. Those are the work orders that make the operation look busier than it is productive.
It saves operating cost. We estimate Oakland's illegal-dumping program runs on the order of $12 million a year in man and equipment time. At that scale, every 1% of crew time moved from unproductive to productive is worth about $120,000 a year. That is the arithmetic behind the break-even in What a Dumpsite Contract Is Really Worth: a contract at roughly $960,000 a year pays for itself when about 8% of crew time becomes productive, which is about 38 minutes per crew per eight-hour shift.
That is the Chicago move. Not more crews, not more trucks, not a new software platform the city has to maintain. Better information about what is actually on the ground, delivered to the person making the dispatch decision, changing which pile the crew goes to first.
The honest caveats
Two things about this list are worth stating plainly, because the alternative is overselling it.
Not every efficiency program saves money. Several well-known municipal technology programs have savings claims that rest on vendor case studies with no published city figure behind them. We checked the smart-waste fill-level sensor category specifically, because it looks like a close fit for the illegal-dumping story, and the available savings numbers are almost entirely vendor-published with no city or independent audit attached. We left those out. If a number cannot be traced to a primary document, it does not belong in a budget conversation, and it does not belong here.
Visibility is not savings. Baltimore's $66 million is exposed spending, not money saved, and we have not counted it as a saving anywhere in this piece. The savings in that story came after, from digital controls, a dashboard, third-party off-duty management, and regular internal audits. Visibility is a precondition. It is worth a great deal to have it, and it is not the same thing as a dollar recovered.
ILLDUMP. Houston METRONow was evaluated and excluded because the published figures describe planned spending rather than verified efficiency savings.
The point
These cities did not cut services. They cut the cost of delivering the same service, by knowing more before they spent.
Chicago reordered its inspections and found the same problems sooner. Los Angeles changed the hardware and paid less to light the streets for twenty years. Pittsburgh got more out of signals that were already installed. Philadelphia stopped sending people to read meters. Franklin planned routes around reality. Baltimore made spending visible and then controlled it.
Illegal dumping works the same way. The trucks exist. The crews are hired. The overtime is budgeted. The gap is that the city is running all of it on a list of guesses about where the piles are, and a large share of that capacity is driving to empty lots and coming back with nothing.
Aerbits supplies the missing measurement. The city gets a verified work queue, the crews get pointed at real piles, and the operation gets faster and cheaper without buying a single additional truck.
The cities in this article all found the money somewhere. We would like to show you where it is in yours.