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.

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 |
How the Engine Works
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