RAAHAT AI: Navigate Smarter, Travel Safer — Building Climate-Aware Navigation with Web Technologies and AWS
How we’re exploring the future of safer, smarter, and environmentally conscious navigation through technology.
I am Passionate Full Stack Developer
Introduction
What if navigation could do more than simply find the fastest route?
Imagine travelling to your destination during heavy rainfall. Your navigation app recommends the quickest road, but that road may be waterlogged. Another route might take a little longer while offering better environmental conditions or fewer reported hazards.
This is the problem we wanted to explore with RAAHAT AI — Navigate Smarter. Travel Safer.
RAAHAT AI is a climate-aware navigation project designed to help people compare routes using available environmental information, travel time, and reported road hazards. Our goal is to make travel decisions more informed by considering not just speed, but also safety and environmental conditions.
- The Problem: Faster Does Not Always Mean Better
Navigation applications have transformed the way we travel. However, choosing a route can become more complicated when environmental conditions change.
Some of the challenges we want to address include:
- Waterlogging and heavy rainfall: Roads may become difficult or unsafe to use.
- Air pollution: Travellers may want to consider air-quality conditions when comparing routes.
- Unexpected road hazards: Reported incidents or blocked roads can affect travel decisions.
- Limited environmental context: Travellers need clear information about which conditions are known and which remain uncertain.
Our idea is to bring these considerations into a single route-comparison experience.
- Our Solution: Introducing RAAHAT AI
RAAHAT AI aims to help users compare different travel options through three recommendation modes:
- Fastest Route: Prioritises travel time.
- Safety-Focused Route: Prioritises available hazard and environmental risk information.
- Balanced Route: Considers travel time alongside available risk information.
The application combines map-based route information with environmental context and a risk-scoring approach. Users can explore routes, compare their trade-offs, and make a more informed choice.
An important principle behind our project is transparency. If live flood data is unavailable, the application should say so instead of presenting an unverified road as safe. Similarly, demo data must be identified as simulated rather than real-world measurements.
RAAHAT AI is an experimental decision-support tool, not a guarantee of road safety.
- Technology Stack: The Tools Behind the Project
We chose technologies that allowed us to build an interactive prototype while learning how mapping, APIs, and cloud architecture fit together.
Frontend Development
HTML5, CSS3, and Vanilla JavaScript
We used HTML to structure the interface, CSS to style the application, and JavaScript to implement the core interactions and application logic.
This approach helped us understand browser-based development without introducing an additional frontend framework.
Mapping and Routing
Leaflet.js and OpenStreetMap
Leaflet powers the interactive map, while OpenStreetMap provides map data.
Nominatim
We use Nominatim for location search and geocoding, helping users find places by name or address.
OSRM — Open Source Routing Machine
OSRM calculates driving routes between locations. These routes provide the foundation for our route-comparison functionality.
Environmental Data
Open-Meteo APIs
We use the available Open-Meteo weather and air-quality APIs to retrieve relevant environmental information.
The exact meaning of each value depends on the data source: some information may represent forecasts or model-based estimates rather than direct measurements at a particular road segment.
Risk Scoring and Demo Mode
JavaScript implements our application logic, including route comparison and risk scoring.
We also created a demo mode with deterministic sample values so we can demonstrate the application when external services are unavailable.
This taught us an important lesson: a reliable demonstration should remain usable during API failures, while clearly distinguishing simulated data from live information.
- Understanding the Architecture
RAAHAT AI currently uses a frontend-oriented architecture. The browser runs the interface, interacts with the map and external APIs, and performs the application logic implemented in JavaScript.
Our planned cloud architecture introduces AWS services to help the project grow beyond a local prototype.
The proposed flow is:
Web Application → API Gateway → AWS Lambda → Data Storage and External APIs → Route Comparison Results
Here is how the main AWS services fit into this design.
Amazon S3 — Static Website Hosting
Amazon S3 can host our HTML, CSS, JavaScript, and other static assets. Amazon CloudFront can optionally improve delivery performance and support HTTPS when configured appropriately.
Amazon API Gateway — Backend Connectivity
API Gateway can expose backend endpoints that the frontend uses to request risk calculations, retrieve hazard reports, or submit new reports.
AWS Lambda — Serverless Processing
Lambda can execute backend code without requiring us to manage a continuously running server.
For RAAHAT AI, it could process environmental information, validate hazard reports, and calculate route-risk scores.
Amazon DynamoDB — Centralised Hazard Storage
Our current community hazard reports are stored in the browser using local storage. That means reports are not automatically shared between different users or devices.
DynamoDB could provide centralised storage so authorised users can submit and retrieve reports through a backend API.
Amazon EventBridge — Scheduled Updates
EventBridge could trigger scheduled backend jobs to refresh selected data sources or run periodic processing tasks. It would coordinate the schedule, not supply weather or flood information itself.
Amazon SageMaker AI — Future Predictive Capabilities
In a future phase, we could investigate training a machine-learning model to estimate waterlogging or other road risks using suitable historical weather, rainfall, road, and incident data.
Such a model would require appropriate training data, validation, and testing before its predictions could be trusted. Predictive machine learning is a future direction, not a capability we claim without implementation.
Architecture note: These AWS services describe our proposed expansion unless explicitly deployed and tested. Keeping that distinction clear is an important part of responsible engineering.
- How Could Route Risk Be Evaluated?
Travel time alone does not describe every aspect of a route.
A risk-scoring system could consider factors such as:
- Available flood or waterlogging information.
- Air-quality conditions.
- Weather conditions.
- Verified or community-reported road hazards.
- The freshness and confidence of the underlying data.
A future implementation could normalise these factors and combine them using documented weights to produce a route-level score.
However, a score is only as reliable as its inputs. Missing flood data, for example, must not be interpreted as evidence that a road is flood-free.
Likewise, a severe verified hazard may need to be treated as a route restriction rather than merely a small penalty in a numerical score.
The objective is to provide understandable comparisons, not to create a misleading impression of certainty.
- Challenges and Lessons Learned
Building this project has highlighted several important engineering lessons.
Working with external APIs: Third-party services may be slow, unavailable, or subject to usage limits. Error handling and a predictable demo workflow are essential.
Handling incomplete data: Different data providers offer different levels of coverage and freshness. Communicating uncertainty is as important as displaying results.
Separating the frontend from the backend: A browser-based prototype is useful for experimentation, but shared data, access control, and more complex processing benefit from a dedicated backend.
Designing for real users: An interface should help users understand the difference between the fastest route and a route that may better fit their preferences. It should also explain limitations clearly.
Planning for scale: Cloud services can support centralised storage and serverless processing, but they also introduce deployment, security, monitoring, and cost considerations.
These lessons helped us look beyond the initial interface and think about the complete system.
- Future Scope
RAAHAT AI has several potential development directions:
- Cloud integration: Deploy the frontend and connect it to a secure AWS backend.
- Shared hazard reports: Allow multiple users to access validated community reports.
- Reliable flood-data integration: Connect a suitable authoritative source where available.
- Improved route scoring: Document and validate how each environmental factor affects recommendations.
- Predictive risk modelling: Investigate machine learning using relevant historical data.
- Notifications: Explore alerts for significant changes in conditions along saved routes.
- Accessibility and usability: Make the application easier to use for different travel needs.
Each feature would need testing against real data and realistic scenarios before being used for safety-critical decisions.
- What This Project Means to Us
RAAHAT AI is more than an attempt to build another navigation interface. It is an opportunity to explore how web development, mapping technologies, environmental data, and cloud computing can work together to address a meaningful problem.
We are learning how to turn an idea into a working prototype, integrate external services, handle failure gracefully, and design an architecture that can evolve over time.
We are also learning that innovation is not only about adding more features. It is about understanding the limitations of the system, communicating them honestly, and building trust through responsible design.
Conclusion
The future of navigation can be about more than reaching a destination quickly. It can also help people understand the conditions they may encounter along the way.
With RAAHAT AI, we are exploring how route comparison, environmental information, community reports, and cloud technologies could contribute to more informed travel decisions.
Our journey is still evolving, and there is more to build, test, and learn.
RAAHAT AI — Navigate Smarter. Travel Safer.
Because a better journey is not only about how quickly we arrive, but also about how thoughtfully we choose our way.
Technologies and References
- OpenStreetMap: https://www.openstreetmap.org/
- Leaflet: https://leafletjs.com/
- Nominatim: https://nominatim.openstreetmap.org/
- Open-Meteo Weather API: https://open-meteo.com/en/docs
- Open-Meteo Air Quality API: https://open-meteo.com/en/docs/air-quality-api
- Amazon S3: https://aws.amazon.com/s3/
- AWS Lambda: https://aws.amazon.com/lambda/
- Amazon API Gateway: https://aws.amazon.com/api-gateway/
- Amazon DynamoDB: https://aws.amazon.com/dynamodb/
#RAAHATAI #AWS #CloudComputing #WebDevelopment #HackathonProject #OpenStreetMap #ClimateTech #LearningInPublic
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article