Skip to main content

AI‑Powered Smart Wayfinding

Mapiddiction’s navigation engine, powered by AI and machine learning, delivers more than just directions — it creates intelligent, customisable routes that adapt to real-world conditions, offering smarter guidance within private maps.

Get Started
*Get started for free. No CC Required.
home page custom map software wayfinding creator

Why Machine‑Learning Routing Beats Traditional Shortest‑Path

Traditional routing ML/AI‑powered routing
Static edge weights set by humans Dynamic weights predicted from real usage, time, density and feedback
Same route for every traveller Personalised to mobility level, transport mode and (soon) user role
Requires manual tweaking to add new rules Continuously learns from new data—no redeploy needed
No awareness of time‑based closures or crowding Context tags (e.g. “night”, “peak”, “service‑only”) steer users automatically
Limited insight for admins Feedback analytics flag bottlenecks and dead zones

Faster & safer journeys

Avoids stairs for wheelchair users, prefers well‑lit paths at night.

Lower maintenance

The rule engine + Random Forest weight predictor tune themselves from data.

Future‑proof

New transport modes or map constraints drop in as just another feature column.

Actionable insights

Heat‑maps and satisfaction scores surface where your map needs attention.

How the Engine Works

1

Rule Engine

Checks obvious constraints first (e.g. "handicap → no stairs").
2

Weight Predictor

Blends 70% baseline weight with 30% ML weight to finesse route choices.
3

Training Pipeline

Retrains on fresh route logs, feedback & context every night.
4

Evaluation System

Tracks accuracy, user satisfaction and API latency so we only ship improvements.

Roadmap

Phase 0 – Initial Launch Available Now

  • Modes: Walking, Bike, Drive, Disabled
  • Base ML engine + ruleset active
  • API integration ready

Phase 1 – Context Tags Coming Soon

  • Time‑of‑day routing (day/night)
  • Density‑aware detours
  • “Avoid stairs” for accessibility

Phase 2 – Preference Profiles Coming Soon

  • Scenic vs Shortest toggle
  • Lift‑first preference
  • Advanced Protected zone avoidance

Phase 3 – Continuous Learning Planned

  • Nightly model retraining
  • Feedback UI integration
  • Quality dashboard

Future Phases Planned

  • Real‑time obstacle feeds
  • Public transport integration
  • Voice guidance & multilingual support
  • Predictive intent routing