Introduction

A hypothetical operator presses Tab and knows that the active control has changed, but cannot see which control will respond to Enter. The last hovered tile is still bright because the pointer remains above it. That highlight is not useful evidence of the keyboard location. Procurement needs a keyboard demonstration that identifies the focused component at each step.

W3C WCAG 2.2 Success Criterion 2.4.7, Focus Visible, at Level AA, requires any keyboard-operable interface to have a mode of operation where the keyboard focus indicator is visible. The Understanding document explains that indicators can be shown only for keyboard use, depending on the user agent and implementation. At least one usable mode must expose focus.

This guide concerns delivered web interfaces associated with a vending project. It does not certify a machine, establish legal obligations or report independent installed-software tests. The three public products below support a retail-format shortlist; keyboard behaviour must be demonstrated separately for the ordered build.

Quick Answer

Ask the supplier to demonstrate a keyboard operating mode, move through the agreed controls and show a visible indication of the current focus at every relevant stop. Then leave focus on a control and confirm that the indicator does not disappear because a timer expires.

The Understanding source states that the focus indicator must not be time limited: when focus is shown, it must remain visible. A momentary animation that fades while keyboard focus stays on the element cannot establish that condition. Record persistence separately from the initial appearance.

SC 2.4.7 does not prescribe a particular indicator shape. Contrast and obstruction have related but separate requirements; Focus Appearance at Level AAA addresses further size and shape questions. Do not call a visibility-only walkthrough a complete accessibility assessment.

Comparison Table

All observations are proposed controlled-test fixtures. None is presented as an installed-product result.

Who Should Buy This

Use this scope when supplied customer or operator web content is keyboard operable. Possible quoted assets include management views, reference portals or web-based configuration tools. Confirm that these assets exist in the quotation before writing acceptance conditions for them.

A touchscreen does not establish keyboard support, and a physical keyboard port does not prove that supplied web content has a visible focus mode. Request the supported input arrangements and the specific interface that will be reviewed. Keep hardware access and web-content behaviour in separate quotation rows.

The scope is especially useful when a supplier replaces browser defaults with custom themes or components. Those changes may affect focus presentation. This is a reason to request evidence, not a claim that custom styling is defective or that any listed WEIMI product removes browser outlines.

How We Evaluate Smart Vending Machines

The shortlist uses WEIMI public product descriptions reviewed on 10 October 2026. We have not performed independent keyboard tests on the installed software of these formats. Published screen, cooling and merchandising facts do not establish a focus indicator or keyboard-operating mode.

The proposed evidence register records build identifier, supported browser or user agent, input arrangement, page and component. It describes how keyboard focus reaches the component, what visibly indicates that focus and whether the indication remains while focus stays there. Separate observations from supplier explanations.

A keyboard walkthrough should use the component’s documented keys. Ordinary sequential navigation may use Tab; a composite menu may use arrow keys once entered. Do not judge a component from a mouse click alone or assume every item must be reached through the same key.

G195 describes an author-supplied indicator as one possible technique. Its examples include outlines and changes to the focused menu item. Techniques can exceed the minimum criterion and are not mandatory designs. Preserve defaults when they provide appropriate evidence rather than assuming every project needs a custom ring.

Key Buying Factors

Identify the keyboard mode. Ask how a user enters the mode in which focus is visible. Browser behaviour can distinguish keyboard from pointer input. A mouse screenshot without a ring does not by itself prove failure; a keyboard screenshot without visible focus remains an unresolved observation requiring investigation.

Separate hover from focus. Place the pointer away from the reviewed component before the keyboard walkthrough. A hover colour can remain on one tile while focus moves elsewhere. Record the component that will receive keyboard activation, using read-only or fictional test actions in an authorised environment.

Check persistence. If an indicator appears as a pulse, allow the pulse to settle while focus stays on the same item. Retain evidence of the resulting state. A decorative animation is not a substitute for a remaining visible indication. Choose an agreed observation window without presenting that duration as a WCAG-prescribed timer.

Include different controls. Review links, buttons, fields and custom components included in the agreed interface. A successful button demonstration cannot establish coverage for a menu implemented differently. A field may show a caret or highlighted text; evaluate the actual visible cue in context.

Review backgrounds and themes. An indicator visible on one panel can become difficult to distinguish on another background. Record supplied themes and states. Related non-text contrast evidence needs its own analysis; this article does not infer contrast ratios from a screenshot.

Distinguish selection. A selected item may remain highlighted after focus moves. Explain how the user can identify current keyboard focus among selected or active states. Do not require a universal appearance, but avoid accepting an unrelated state marker as proof of focus.

Handle dynamic transitions separately. When an overlay opens, determine where focus goes and observe its indicator there. Focus order, trapping, return location and obstruction require separate checks. Visibility evidence records the current position without closing those broader findings.

Retain build-specific evidence. A component library may document an indicator that a delivered theme overrides. Request the final build, not only an upstream sample. Recheck relevant components after theme changes or replacement controls.

Best Smart Vending Machines

“Best” means a fit between the public merchandising format and the buyer’s plan. These are not independently tested keyboard or accessibility rankings.

Feature Comparison

Open evidence means that a listing cannot establish the requested behaviour. It does not mean that a product failed an independent test.

Cost & ROI Analysis

This illustrative USD scenario budgets a bounded review using assumed effort and rates. It is not a supplier quote, measured saving or prediction of defects.

The planned review costs $270 less than one assumed correction scenario. Break-even is $2,100 divided by $2,370, approximately 0.89 equivalent correction events. No avoided event is asserted. Replace assumptions with actual supplier estimates before deciding scope.

Hardware, merchandise revenue, payment fees and a full accessibility assessment are excluded. Price related contrast, obstruction or keyboard-navigation work separately if it is outside the agreed walkthrough. Do not claim that this limited budget establishes complete conformance.

Best Choice by Scenario

Packaged-product retail: Shortlist the AI vision fridge when its public merchandise description fits. Confirm access, payment and any associated web content. Camera checkout does not establish keyboard navigation or a focus indicator.

Touchscreen dispensing: Consider WM22 after selecting the mechanism and supported software scope. Ask whether keyboard operation is delivered for the relevant web assets, then request visible-focus evidence for that build.

Two selling areas: Consider the dual-cabinet station when its physical arrangement matches the site. Do not infer keyboard behaviour from the main display or cabinet count. Request actual interface boundaries and input arrangements.

Existing operator portal: If equipment will use a separately supplied portal, assign responsibility for its keyboard evidence to the appropriate software owner. Avoid attributing portal behaviour to the hardware listing.

Applications

A hypothetical reference portal can demonstrate keyboard focus across navigation links and expandable help controls. Record each current location before activation. The fact that a link opens the correct page establishes its destination, not that focus was visible before activation.

A hypothetical management menu can use arrow-key movement within a composite component. Capture the focused item after the movement stops and distinguish it from any persistent selection. The proposed fixture is not evidence that a shortlisted product uses this menu design.

A theme acceptance session can compare light and dark panels with the same component focused. Store the supported configuration and the observation. Do not claim a contrast measurement until the relevant colours, states and method are assessed.

A release that introduces new controls should add them to the component register. A repeated screenshot of the old button does not establish coverage of the new control. Keep initial appearance, persistence and unresolved related checks visible in the handover.

FAQ

Final Recommendation

Choose the retail format from verified product facts, then agree on a supported keyboard mode and a component-level walkthrough for supplied web content. Retain the visible focused state and persistence evidence for the delivered build. Make responsibility clear when third-party software is involved.

Keep hover, selection and keyboard focus distinct. Review the current indication without imposing a universal shape, and keep contrast, obstruction, ordering and keyboard-exit findings in their own scopes. The result is a reviewable procurement requirement rather than an unsupported claim about machine accessibility or business performance.

Sources: W3C Understanding SC 2.4.7 and G195, read on 10 October 2026. Product links support public retail-format facts only. Interface fixtures and finance are hypothetical; no certification, indexing, demand or SEO outcome is claimed.

CTA