San Diego City Council’s agenda for Tuesday 25 August 2026 carried, at Item 58, a sixth amendment to the city’s container contract with Rehrig Pacific Company. According to the San Diego Community Coalition’s weekly bulletin, published by OB Rag, that amendment takes the running total for trash and recycling container replacement and repair to $106.5 million, against $8.5 million for the first amendment back in 2020. The same bulletin adds a line that anyone who has watched a municipal RFID programme scale will recognise: “It’s still unclear how data from the RFID (radio identification) chips are used.” Those figures and that observation are the coalition’s, not the city’s. No outlet has reported how Council voted, and we have not seen the city’s staff report for the item, so the outcome is still open.
The scale underneath that agenda line is the reason it matters. By 24 March 2026 the City of San Diego Environmental Services Department had delivered 231,178 new gray trash bins to 215,610 customers and removed 257,508 old black ones, covering 96% of eligible customers against a base of more than 225,000 households. Every one of those containers carries what the city calls a scannable reader tag. That makes this one of the largest municipal cart tagging exercises in the United States. And yet the part that turns a tagged bin into operational data, the reader on the truck, is still described by the city itself as a pilot.
Billing came first, route optimisation second
For more than a century San Diego collected residential waste at no direct charge, under the People’s Ordinance. Voters passed Measure B in 2022, which opened the door to charging, and in June 2025 Council approved a Solid Waste Management Fee of $523.20 a year, billed on county property tax bills in two instalments of $261.60.
That single change explains the entire technology decision. To bill a property for a service, you have to know which container belongs to which property, and you have to be able to prove it at the kerb. The city’s FAQ is unusually direct about this: the gray trash bins are “City-provided” and carry “unique serial numbers assigned to a specific property”, and from 1 July 2026 “only official City-provided gray trash bins are eligible for trash collection.” The container swap is, first and foremost, an eligibility control. It also did some sorting on the way through: 95% of newly ineligible households had moved to private haulers by late March.
The billing structure reinforces the point. Default service is one 95-gallon (about 360 litres) trash container and one 95-gallon recycling container. Smaller 35-gallon (about 130 litres) and 65-gallon (about 245 litres) recycling and organics containers are available for space-constrained properties “at the same rate”, with billing pegged at the 95-gallon level. The charge does not vary with volume. What the city needs from the container is identity, not capacity. That is a meaningfully different starting point from most RFID waste bin deployments, which are usually justified on diversion rates or pay-as-you-throw.
The deployment, dated
Gray bin deliveries began on Monday 13 October 2025 and were scheduled to run into spring 2026. At the start of the programme roughly 300,000 trash cans were in use across the city, three-quarters of them more than 20 years old, so this was a fleet at end of life regardless of the billing change.
By 24 March 2026 the gray phase was effectively done, and the light blue recycling bin phase began the same day, running Monday to Friday citywide with some properties not receiving containers until early fall 2026. Across both phases more than 750,000 old bins are being collected and recycled.
Jeremy Bauer, Assistant Director for Environmental Services, said the new bins feature “scannable tags that will help the City track performance and continually improve service.” One date worth holding on to: weekly recycling collection starts on 1 July 2027, replacing the current fortnightly service. A frequency change is exactly the point at which service verification data stops being nice to have, because complaint volumes spike whenever residents have to relearn a schedule.
What is actually in the bin, honestly
Kirby Brady of the Environmental Services Department described the hardware plainly in October 2025: the new bins have “Radio Frequency Identification (RFID) chips hidden in the handle.” The city’s position is that those chips “will make it possible for trash trucks to scan and confirm the can is at the correct house.” The FAQ describes the same component as “a scannable reader tag” and says the city “is piloting the use of that technology to track when our trucks empty a container, helping us increase effectiveness, efficiency, and accountability.”
That is the whole of the published technical record. San Diego has not stated the air interface, the frequency band, the tag model, the memory layout, whether the identifier in the tag matches the serial number printed on the container, or the encoding scheme. Anyone telling you these are RAIN RFID tags is inferring, not reporting.
The inference is a reasonable one, and it is worth setting out as general industry context rather than as fact about San Diego. Kerbside cart tagging in North America is overwhelmingly passive UHF in the 902 to 928 MHz band, typically to the ISO 18000-63 air interface, because that combination gives a few metres of read range, tags cheap enough to mould into every container whether or not the customer ever switches the system on, and a mature supply of truck-mounted reader hardware. The older approach, low frequency tags read at a few centimetres by an antenna in the lifter grabber, is very hard to mis-read but requires the tag and antenna to be almost touching. Rehrig Pacific moulds RFID into carts at the point of manufacture precisely so the container can carry the capability regardless of when the buyer activates it. None of which tells you what San Diego actually specified, and the city has not said.
Read geometry, and why the missed read is the hard problem
The difficult engineering problem in municipal cart tagging is not tag cost, which is trivial at manufacture, and it is not database scale. It is the read window. With the tag in the handle and the antenna on the arm or body of the truck, the system has the second or two in which the grabber closes, lifts, tips and returns to capture a read. On an automated side loader that geometry is repeatable. On a rear loader with manual presentation it is not.
The failure modes are mundane and relentless: a resident turns the bin so the tag faces away, a container is presented tight against a parked car and the arm reaches at an angle, rain and nearby metalwork detune the antenna, or two containers sit close enough that a generous read zone picks up the neighbour’s tag instead. At a 100% read rate you get proof of service. At 97% you get 3% of collections that look like non-service, and across hundreds of thousands of lifts a week that is thousands of false exceptions a day, every one of which either gets adjudicated by a person or quietly erodes trust in the data.
That is why cities pilot the truck side separately, and San Diego’s sequencing, tag every container at manufacture and prove the reading layer afterwards, is defensible engineering rather than a hedge.
Nobody has named who reads the tags
Rehrig Pacific is confirmed as the container manufacturer under a $64.7 million contract approved by Council in July 2025, split $41.5 million for the new bins and $23.2 million for delivery plus collecting and recycling the old ones. Rehrig has recycled damaged bins for the city since 2018, so the relationship long predates this programme.
What is not confirmed anywhere is who supplies the truck-mounted readers, who integrates them with the fleet telematics, or where the reads land. Rehrig sells its own answer to that question: a service verification product built on RFID readers fitted to collection vehicles, feeding its Vision software, giving real-time confirmation that a container was serviced, cart inventory and location, and a route-side mechanism for drivers to flag contaminated carts. Rehrig’s published service verification customer references name Solterra and Taunton, Massachusetts. They do not name San Diego.
The honest reading is that San Diego has bought RFID-ready containers from a supplier that also sells the reading layer, and has not yet committed to that layer. Whether the reader side is inside the sixth amendment or arrives as a separate award is one of the more interesting things to watch here.
What service verification buys when it goes live
Four things, in roughly descending order of value to a city this size.
Proof of service against missed-collection complaints. In most kerbside operations this is the largest single category of resident contact, and the expensive part is not the call, it is the return truck dispatched on the resident’s word alone.
Contamination flagging on route, where the driver marks the container at the moment of the lift rather than through a downstream audit that cannot be tied back to an address.
Cart inventory and location, so the city knows what asset sits where without a manual survey. Against 750,000 containers changing hands in under a year, that is not a small administrative saving.
Stray and unauthorised cart recovery, which becomes materially more important after 1 July 2026, when non-city containers stop being eligible for collection.
The caveat matters: the published performance numbers for all of this belong to Rehrig’s Vision customers, not to San Diego, and San Diego has published neither a baseline nor a target.
The circular story is about the old bins, not the tagged ones
The end-of-life side of this programme is genuinely strong and deserves to be reported on its own terms. Old containers are collected at the kerb, taken to a Miramar city yard and cleaned. Wheels and metal bars go to regional recyclers. The remaining plastic goes through a universal refiner that chips the bins into smaller pieces, and crucially that chipping happens on site at the Environmental Services Collection Services yard. Andrea Deleon, Deputy Director of ESD’s Waste Reduction Division, said the on-site approach “reduces the number of trips required and lowering overall carbon footprint” (quoted as printed).
The chips go to Rehrig Pacific plants in Los Angeles and Arizona and re-enter production as brand-new bins or other reusable transport packaging including totes, pallets and trays. Remaining material goes to toll grinders in California to become pails, composite railroad ties and conduit fittings. More than 60,000 old bins had been collected by late November 2025, against roughly 750,000 to be processed by summer 2026.
None of that is tag-enabled. The old bins have no chips in them, and it would be an overclaim to suggest RFID is doing any work in that loop. What the tags change is the next cycle. For the first time the city holds a per-asset record for every container it owns, bound to a property, from the moment it entered service. That is the raw material for warranty claims, failure-rate analysis by production batch, and a properly evidenced replacement decision in 20 years’ time rather than another citywide guess. It is the quietest long-term benefit of municipal cart tagging and it has gone almost entirely unmentioned in the coverage.
The questions still open
What is the data actually used for? The coalition bulletin’s complaint is the right one. Billing verification and service verification are different systems with different retention profiles and different audiences inside a city, and San Diego has not separated them in public.
Who reads the tags, on what hardware, under what contract? Still unnamed, in a programme that has now been running at scale for ten months.
What were residents told? The coverage from October 2025 onward carries no privacy or data-use discussion at all. A chip hidden in the handle, bound to a specific property, read by a passing vehicle, is functionally a per-household identifier at the kerb, and set-out data accumulated over time is a behavioural record. Cities that have run into difficulty on this have generally run into it after the pilot scaled, not before. Retention, access and whether reads ever feed enforcement are cheaper to answer now than after the readers are on every truck.
And what happens when the pilot does scale? Someone has to write down an acceptable read rate, an exception handling policy and who carries the cost when a lift is not captured. On the evidence available, that sentence is not yet in public.
Item 58’s outcome will tell us how much more money is going into containers. It will not, on its own, answer any of the above.
