Palm Logo
July 3, 2026·PalmAI-ProductTeam

Autonomous Stores and Cashierless Checkout: How the Technology Works

TL;DR

An autonomous store is a retail format where customers pick up items and leave without a manual checkout, because a combination of computer vision, sensors, and a linked payment identity records what they take and charges them automatically. Cashierless checkout is the specific mechanism that makes this possible — "just walk out" shopping. In 2026, unmanned and frictionless retail is one of the most-watched retail trends, with new deployments from convenience chains to fully robot-run stores. This guide explains how the technology actually works — sensing, vision, and identity — and why the identity-and-payment layer, increasingly biometric, is the piece that makes "walk out and go" trustworthy.

Who This Article Is For

  • Retail innovation and store-format leaders evaluating autonomous concepts
  • Payment and technology product managers designing the identity/payment layer
  • Systems integrators assembling the sensing and checkout stack
  • Anyone researching how cashierless and unmanned stores work

What Is an Autonomous Store?

An autonomous store — also called an unmanned, cashierless, or checkout-free store — is a retail format where the act of checking out disappears. Instead of scanning items at a register, the customer takes what they want and leaves; the store's systems detect the selections and charge a pre-linked payment identity automatically.

The category spans a spectrum. At one end are staffless convenience formats with a self-checkout kiosk; in the middle are "just walk out" stores using cameras and shelf sensors; and at the leading edge are fully robot-run stores, including a recently opened autonomous, robot-operated retail store in Shanghai powered by embodied AI. What unites them is the removal of the manual checkout step.

How Cashierless Checkout Works

A "just walk out" experience is the visible tip of a coordinated technology stack:

  1. Entry and identification. The customer identifies themselves on entry — historically by scanning an app QR code or tapping a card — which opens a session tied to a payment method.
  2. Sensor fusion. Overhead cameras, shelf weight sensors, and sometimes RFID track which items are picked up, put back, or carried out.
  3. Computer vision. AI models associate each item with the right shopper's virtual cart in real time, handling the hard cases — a customer picking up two items, then returning one.
  4. Exit and auto-charge. When the customer leaves, the session closes and the linked payment identity is charged automatically; a digital receipt follows.

The engineering difficulty is concentrated in steps 1 and 3: correctly attributing every item to the right person, and reliably linking that person to a valid means of payment. Sensing has advanced quickly. The weaker link, historically, has been the identity step — which is where the format most often adds friction back in.

The Real Bottleneck: Trustworthy Identity

The paradox of frictionless retail is that it still has to solve one high-friction problem up front: reliably knowing who is shopping and whose account to charge. Early "walk out" formats required a smartphone app, an account, and a QR scan at the gate — which reintroduces exactly the friction the format promises to remove, and excludes anyone without the right app installed.

This is why the identity-and-payment layer is increasingly treated as the make-or-break component of an autonomous store, not an afterthought bolted onto the sensing stack. If entry and payment can be collapsed into a single, natural action that anyone can perform without a phone or app, the format finally delivers on "frictionless." That requirement is precisely what points toward biometrics.

Where Biometric Identity Ties the Store Together

If an autonomous store needs to know who is shopping without demanding a phone and an app, a biometric that is fast, contactless, and hard to spoof becomes the natural entry-and-payment key. Palm recognition — a contactless biometric method that identifies a person from their palm print and the vein pattern beneath the skin — fits this role well: the customer presents an open hand at entry, the session binds to their enrolled identity and payment method, and the same gesture can authorize the exit charge.

In Tencent PalmAI's PayMax retail deployments, identity and payment collapse into one action — the hand that identifies the shopper is the hand that pays — which is exactly the property an autonomous store needs to remove the app-and-QR step at the gate. When Tencent PalmAI's palm payment launched with 7-Eleven in China, it reached 1,500 stores within a month, and store operators reported roughly a 25% improvement in cashier efficiency at high-traffic checkouts — a signal of how much the identity-and-payment step drives throughput in unmanned formats.

Palm does not build the store's vision system or shelf sensors. It solves the one problem those systems cannot solve on their own: binding the shopper to a trusted identity and payment method with a single, phone-free gesture.

Where Palm Fits in an Autonomous Store

  • Frictionless entry and payment. One open-hand gesture opens the session and authorizes the charge — no app required. This is the Retail & Payment pattern PayMax targets.
  • Inclusive access. Because it needs no smartphone, palm entry extends autonomous retail to customers an app-gated store would exclude.
  • Cross-format identity. The same enrolled palm can work across a chain's staffed, self-service, and unmanned locations.

Limitations and Considerations

Autonomous retail is advancing fast, but honest evaluation matters:

  • Sensing accuracy and shrink. Item-attribution errors and loss remain active operational challenges independent of the payment method.
  • Palm needs dedicated hardware. A palm reader is a purpose-built sensor at entry or checkout, not a repurposed store camera.
  • Enrollment is one-time and in person. Customers register their palm once at a sensor; there is no fully remote enrollment.
  • Ecosystem and readiness vary. Fully autonomous formats scale best where a supporting payment ecosystem exists; in other markets, biometric entry is additive to app- and card-based access.
  • Compliance mapping. Palm data is biometric data under GDPR, PIPL, LGPD, and similar frameworks; consent, retention, and necessity should be reviewed with a data protection officer before deployment.

Frequently Asked Questions

What is the difference between an autonomous store and a self-checkout store?

A self-checkout store still has a checkout step — the customer scans and pays at a kiosk. An autonomous ("just walk out") store removes the checkout entirely: sensors and computer vision detect what is taken, and a linked identity is charged automatically on exit.

How do cashierless stores know what I took?

They fuse multiple signals — overhead cameras, shelf weight sensors, and sometimes RFID — and use computer-vision models to associate each item with the correct shopper's virtual cart in real time, including when items are picked up and put back.

Why do many autonomous stores still require an app?

Because the hardest part is reliably identifying the shopper and linking a valid payment method. App-and-QR entry was the early solution, but it adds friction and excludes non-app users — which is why biometric entry is emerging as a phone-free alternative.

How does biometric palm entry work in an unmanned store?

The customer enrolls their palm once, then presents an open hand at entry to open a session tied to their identity and payment method; the same biometric can authorize the exit charge. It requires a dedicated reader and consent. Retailers can review the PayMax deployment pattern or use the contact form on this page.

Related Resources

About Tencent PalmAI

Tencent PalmAI is an AI-powered palm recognition service combining palm print and palm vein identification, protected by 90+ patents and validated through 20+ peer-reviewed conference papers. PalmAI products span high-volume payment authentication (PayMax), identity verification (KYCMax), edge access control (SmartLock), and offline enterprise deployment (Standard).

To evaluate palm recognition as the identity-and-payment layer for an autonomous store concept, use the contact form on this page.

→ Learn more at palm.tencent.com

Sources

Ready to start ?
Use PalmAI in your business now!