Roadmap

The Systemic Accessibility Analyzer is evolving from a multi-tool reporting platform into an accessibility intelligence engine.

Rather than simply aggregating reports, the project is focused on turning accessibility evidence into actionable insight.

Don't count findings. Find consensus.

Current focus: Phase 1 has successfully established the robust multi-tool evidence layer, PostgreSQL job data orchestration, engine-family badges, the Fix Once, Benefit Many prioritization engine, user-defined WCAG target levels (A/AA/AAA), and the new Color Vision Deficiency (CVD) simulation tool. Phase 2 will expand deeply into automated component intelligence, deeper confidence weighting, and root-cause mapping.

Current

Phase 1 — Evidence Engine

Collects, normalises, compares, and visualises accessibility evidence from multiple engines with interactive drilldowns and companion visual tools.

Phase 2 — Accessibility Intelligence

Moves from individual findings towards advanced component grouping, confidence scoring, repeated patterns, and root-cause identification.

Future

Phase 3 — Guidance

Explores suggested code repairs, long-term regression trends, design-system insights, and prioritised remediation tracking.

Phase 1 — Accessibility Evidence Engine

Phase 1 focuses on gathering accessibility evidence from multiple independent sources and presenting it in a unified, meaningful way.

Instead of manually comparing reports from numerous accessibility tools, the analyzer normalises findings, removes duplication where possible, and highlights where tools independently agree.

Phase 1 asks: What did the tools find, and where do they agree?

Current capabilities

Support for multiple accessibility engines and companion evidence tools.
Public and authenticated website scanning orchestration.
Adapter-based report ingestion and PostgreSQL framework integration.
WCAG mapping, fallback engines, and metadata enrichment.
Rule canonicalisation across diverse tool families.
Cross-tool deduplication and engine-family categorization.
Consensus analysis and Shared Pattern Impact scoring.
Interactive dashboard views with paginated and searchable text drilldowns.
Static report generation for published job runs.
Rich JSON and multi-sheet Excel workbook exports (Power BI ready).
Visual Explorer with page previews and keyboard Tab Map views.
Virtual Screen Reader simulation streams.
Visual Contrast Checker logs.
HTML validation and Page Inventory coverage checks.
Benchmark website suite.
User-defined WCAG target level filtering (A, AA, AAA) for tailored compliance reporting.
Color Vision Deficiency (CVD) simulation views.

Supported evidence sources

Tool / view	Purpose
axe-core	Automated WCAG testing and widely used accessibility rule coverage.
axe-scan	Axe-derived reporting through an additional scan pathway.
IBM Accessibility Checker	Confirmed, potential, and advisory accessibility findings.
HTML CodeSniffer / HTMLCS	WCAG technique-style rule evidence and review-oriented findings.
Lighthouse	Browser-based accessibility audit metadata and failures.
Oobee	Crawl-style accessibility evidence and axe-family results.
UUV	Flow-based and heuristic accessibility evidence.
Alfa	ACT-style outcomes including failed and needs-review results.
Nu HTML Checker	Markup, validation, and standards-level evidence.
Pa11y Axe / Pa11y HTMLCS	Additional axe and HTMLCS evidence through Pa11y runners.
Visual Explorer	Clean page previews and keyboard tab-order maps from the same capture flow.
Virtual Screen Reader	Structured reading-order and semantic output.
Contrast Checker	Visual contrast evidence and colour-pair analysis.
CVD Simulator	Color Vision Deficiency (colorblindness) simulation to evaluate visual accessibility barriers and color reliance.

Benchmark suite

The analyzer is continually tested against a growing benchmark suite representing different kinds of real-world accessibility evidence.

Benchmark type	Why it matters
Government / reference-quality sites	Useful baseline where stronger accessibility implementation is expected.
Retail / ecommerce	Tests product pages, navigation, imagery, forms, and transactional patterns.
Modern SPA / Next.js	Exercises component-heavy, JavaScript-rendered, creative and animation-rich websites.
Legacy / malformed HTML	Tests old markup patterns, validation failures, frames, and structural oddities.
Deliberately poor UX	Useful for hostile interaction design, keyboard problems, confusing forms, and regression testing.
Authenticated flows	Validates login, storage state, protected-page analysis, and real application journeys.

Phase 2 — Accessibility Intelligence

Phase 2 shifts the focus away from individual findings and towards understanding accessibility at a higher level.

Instead of answering:

How many issues were found?

the analyzer will begin answering:

What needs fixing?

Component analysis

Findings will be associated with likely interface components rather than grouped only by page or WCAG rule.

Navigation
Forms
Buttons
Links
Images
Cards
Tables
Dialogs
Carousels
Headers and footers

The goal is to identify repeated design-system issues rather than isolated page defects.

Root cause analysis

Many accessibility reports contain hundreds of repeated findings. Phase 2 aims to identify the underlying repair.

Instead of	The analyzer identifies
63 button findings	A shared Button component pattern.
41 contrast failures	A shared colour token or visual treatment.
82 missing label findings	A repeated Form component issue.
Many image-alt issues	A repeated Image or Gallery component problem.

Confidence scoring

Not every accessibility finding carries the same weight. Future scoring can consider:

Agreement between tools.
Agreement between independent engine families.
WCAG consistency.
Component similarity.
Shared DOM targets.
Confirmed versus advisory or needs-review findings.

Accessibility heatmaps and visual layers

Future visualisations may include:

Landmark overlays.
Heading hierarchy views.
Form analysis overlays.
Image analysis overlays.
Component boundaries.
Accessibility heatmaps.

These visual layers build naturally on the current Visual Explorer, Virtual Screen Reader, and Contrast Checker outputs.

Phase 3 — Accessibility Guidance

Phase 3 is intentionally more exploratory. The broad idea is to help teams move from understanding evidence to planning repairs.

Potential ideas include:

Suggested repair descriptions.
Design-system ownership hints.
Component-level remediation summaries.
Trend analysis between benchmark runs.
Organisation-wide accessibility intelligence.
Exportable prioritised repair plans.

Scope reminder: The analyzer is not intended to replace manual accessibility expertise. It organises automated and companion-tool evidence so that human review and remediation decisions can be better targeted.

Long-term vision

Accessibility tools already identify problems. The Systemic Accessibility Analyzer is focused on understanding those problems.

Ultimately the goal is to help answer questions such as:

Which issues have the strongest evidence?
Which findings are repeated across components?
Which repairs remove the greatest amount of accessibility debt?
Which issues genuinely require manual review?
Which components should be prioritised first?

From findings to evidence. From evidence to consensus. From consensus to meaningful repairs.

Accessibility Analyzer Roadmap

Roadmap

Phase 1 — Evidence Engine

Phase 2 — Accessibility Intelligence

Phase 3 — Guidance

Phase 1 — Accessibility Evidence Engine

Current capabilities

Supported evidence sources

Benchmark suite

Phase 2 — Accessibility Intelligence

Component analysis

Root cause analysis

Confidence scoring

Accessibility heatmaps and visual layers

Phase 3 — Accessibility Guidance

Long-term vision