Claire Cosson, writing on Zepros, an online journal for the French restaurant and catering industry, has published an interesting article about what happens when a reusable packaging scheme grows up and the losses start to bite. Since 7 September 2026, the stainless steel meal containers used by the home delivery service in Fontenay-sous-Bois have been tracked by RFID tag. Cosson, editor in chief of Zepros La Collective, has covered French collective catering for more than twenty years, and her piece is notable less for the deployment than for what the commune put on the record: the real costs, the readers that did not work, and the residents who will never hand a container back.
Fontenay-sous-Bois is a commune in the Val-de-Marne department (94), in the eastern suburbs of Paris, and its municipal central kitchen delivers meals to residents at home. Reuse there is five years old. The town began trialling stainless steel containers from Cuitisan in September 2021 with 45 elderly recipients, prompted by France’s AGEC law, the anti-waste law for a circular economy adopted in February 2020, which required home meal delivery to stop using single-use plastic containers from 1 January 2022. ADEME, the French environment and energy management agency, put EUR 53,900 towards the infrastructure, including a washing tunnel and vehicle modifications.
Reuse was not the new part
Maxime Cordier, the town’s director of catering and president of Agores, the French national association of directors of municipal collective catering, made that distinction plainly to Zepros. The town had already launched reuse and knew how to run it. What started in September 2026 is reuse with chip tracking.
The driver is the one every returnable asset programme meets eventually. “When you do this without any tracking, at some point the thing goes missing,” Cordier told Zepros. The containers cost EUR 7 for a small one, EUR 9 for a medium and EUR 11 for a large, including tax, against roughly 1,600 in circulation, so losing a few hundred a year quietly rebuilds the cost case that reuse was meant to win. Ninety of the town’s 180 home delivery recipients are inside the tracked scheme so far.
Handhelds on the round, a portal at the wash line
The architecture is the familiar two-point shape for any loop where assets leave a building and come back. On delivery rounds, drivers carry a reader that identifies several containers at once rather than one at a time. “You do not want to be like the door of a theatre, reading them one by one,” Cordier said. On return, two antennas in the wash area form a portal that records containers automatically as they pass through. Between them, the two read points tell the kitchen what went out, to which recipient, and what came back. The same chokepoint shows up at Cheonan’s reusable delivery container centre in South Korea, where RFID logs every wash cycle, and at Mainzer Weinmarkt, where reusable dishes pass through deposit return machines.
Food-contact construction and read range
Reliability was the hard part, and the cause is worth spelling out for anyone about to specify something similar. A tag that has to be food-contact safe and survive a microwave is physically constrained: it must be encapsulated, and the materials and geometry that make it safe also cost it performance. “The tag was not transmitting strongly enough,” Cordier said of the first setup, describing a classic shortfall in read range on passive tags working through an awkward package.
The fix was to change the readers rather than the tags, which is usually the right instinct. Passive tags have no battery, so the energy in the link comes from the reader side. Once a tag is at the limit of what its enclosure allows, you buy sensitivity and antenna quality at the reader, not at the item. On the microwave question, Cordier said the tag is rated microwave-safe and the town ran hundreds of tests on it. The same constraint is shaping tag design elsewhere, as with Tageos and its EOS-450 MW microwave-compatible inlays for food packaging. One caveat: the Zepros article does not state the frequency or air interface the town chose, and we will not guess one. The description implies reading several items at a distance, but the band is not disclosed.
The money sits in the infrastructure
The town put the cost at around EUR 15,000 in the first year including licensing, then EUR 1,000 to EUR 2,000 a year. Cordier’s summary of where it goes is the most useful line in the piece for anyone building a business case: “What costs a lot is not so much the tags themselves, it is the readers and the antennas.” That is the standard shape of an RFID project. Tag price scales with item count so it dominates the conversation, but at 1,600 items the fixed infrastructure is what you are really buying, and the economics improve sharply with volume. Our explainer on RFID in reusable packaging covers that arithmetic.
The 20 per cent no reader can fix
The most honest part of the article has nothing to do with technology. Cordier’s real limit is the recipients: “There are always 20 per cent of the population for whom it does not work,” he said, and for the rest, “for 70 to 80 per cent, it works.” Read that carefully, because it is easy to misreport. Those are not read rates. They are the share of elderly home delivery recipients who reliably put the container back out for collection. RFID tells the kitchen exactly which containers did not come home and who has them, which is valuable, but it does not make a frail resident hand a tray back. Knowing where the losses are is a different thing from preventing them.
Why one commune matters
Because Cordier chairs Agores, his verdict travels further than Fontenay-sous-Bois. French communes are under sustained pressure to move to reusables: AGEC covered home meal delivery from 2022, and article 28 of the EGalim law banned plastic containers for cooking, reheating and serving in school and university catering, and in settings for children under six, from 1 January 2025, with communes under 2,000 inhabitants given until 1 January 2028. Thousands of kitchens are accumulating stainless steel containers, and every one will meet the same losses. Fontenay-sous-Bois has published a working reference point: 1,600 containers, EUR 15,000 to get tracking running, a few thousand a year to keep it, readers rather than tags as the main cost, and a human ceiling on how much leakage you can close.
Read more at https://resto.zepros.fr/restauration-collective/gestion-directe/fontenay-bois-trace-ses-barquettes-rfid
