Why UPI Fraud Is a Temporal Logic Problem (and How PLCs Solved It 50 Years Ago)
UPI fraud detection is probabilistic and post-authorization. The money moves, then a model scores it 0.73 and a fraud analyst looks at it in two days. That's a power plant that checks the boiler after it explodes. In 1981, IEC 61131 defined the PLC. The core principle: a hard fault gets a hard response. No probability. No score. The interlock trips, and the machine stops. Same input → same output…
UPI fraud detection relies on probabilistic and post-authorization methods, with money moving before a model scores its likelihood of fraud. An analyst then reviews the transaction two days later. This approach is likened to a power plant monitoring its boiler after an explosion. In 1981, the International Electrotechnical Commission (IEC) 61131 introduced Programmable Logic Controllers (PLCs).
The core principle of PLCs is straightforward: a hard fault triggers an immediate, irreversible response. There is no probability or scoring involved. The interlock trips, and the machine stops. This principle of "same input → same output → every time → forever" was applied to UPI payment events by the fraud-watchdog library, a Python tool that checks multiple aspects of each payment.
The library examines three axes for every transaction: velocity (transaction rate), dwell times, and sequence. Scenarios include a payment confirmed in four seconds after the arrest call, a sequence required for event order (screen-off → resume without re-authentication), and a time-of-day context (18-second burst at 3 AM outside the user's active window).
The library successfully tripped 6 out of 8 scenarios, with no false positives. It operates on a cost-effective device, requiring only a ₹6,000 phone. The project is community-driven, with users encouraged to submit their own scam scenarios for inclusion.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
