The tiny chip that powers Montreal subway tickets
Montreal’s single‑use subway tickets hide a grain‑of‑salt–sized MIFARE Ultralight NFC chip, prompting a broader look at how modern transit fare systems work technically, economically, and from a user‑experience standpoint. Commenters compare NFC tickets with alternatives like QR codes, magnetic stripes, and contactless bank cards, weighing trade‑offs in security, latency at turnstiles, offline operation, cost, and interoperability with phone wallets. Many see disposable NFC as a practical way to unify fare collection hardware, but debate its environmental impact and argue that reusable cards, open‑loop payments, or even fare‑free transit would be better long‑term directions.
NFC ticket technology & security
- Montreal’s disposable tickets use MIFARE Ultralight EV1, a very simple ISO 14443A chip with a signed UID and limited memory; backend systems typically treat it as a bearer token.
- More secure families (MIFARE DESFire, Ultralight C, Felica, etc.) store cryptographic keys in tamper‑resistant hardware and support mutual authentication and encryption.
- Ultralight has a password mechanism; some argue this meaningfully raises the bar for cloning, others note you can harvest the password mid‑transaction and then clone, so backend checks and duplicate‑use detection remain important.
- EMV contactless payments sit on the same RF stack but are account‑based; transit cards are usually stored‑value, enabling fast offline operation and resilience to network outages.
Manufacturing & chip design
- Chip is built on an older ~180 nm process; the die is grain‑of‑salt sized, with ~45k transistors and an analog front‑end for RF power and load‑modulation.
- Wafers are thinned (back‑ground) to tens of microns, diced with ~20 µm saws, and dies picked with automated handlers down to ~0.2 mm.
- UIDs are programmed and permanently locked during wafer test; other EEPROM areas can be one‑way locked and support anti‑tearing counters via indirection schemes.
- Antennas are typically printed conductive ink or etched metal on plastic; details are mentioned but not deeply explored.
Transit systems, UX, and latency
- Many systems cited: DESFire (Clipper, Oyster), MIFARE Classic/Plus (Boston, Moscow), Felica (Japan, Hong Kong), Calypso (Montreal OPUS), EMV‑based (NYC OMNY, London, Sydney).
- High throughput at gates drives design: RFID/NFC taps (~100–500 ms) are preferred over magstripe and QR, which are seen as slower, less robust, and more angle/lighting‑sensitive.
- QR is viewed as attractive for cost and flexibility, but problematic for offline use, multi‑ride tickets, and fast gates; some cities (e.g., parts of China, India, Japan soon) still pursue QR with heavy backend infrastructure.
- Phones and wearables as transit tokens: Apple uses a secure element and “Express” mode; Android widely supports Host Card Emulation, though without hardware security by default.
Cost, waste, and policy
- Per‑chip prices around a few cents; some argue mechanical mag readers and maintenance cost more than the electronics.
- Debate on e‑waste: some see single‑use ICs and antennas as unjustifiable; others note the mass per ticket is tiny and most frequent riders use reusable cards.
- Privacy concerns around bank‑card tap‑to‑ride and account‑based systems are raised; others point out many systems still allow anonymous stored‑value cards or cash‑bought disposables.
- Several commenters advocate free or flat‑rate public transit to eliminate fare technology complexity and reduce transaction friction entirely.