MechanIQs

Favicon
Sustainability impacts
Why make this?

Inspiration

We realized that Neil Postman's quote about living in a world that makes no sense perfectly describes how most people feel about their cars. Millions of drivers want to maintain their own vehicles but are quickly overwhelmed by confusing parts pricing, a lack of vehicle-specific guidance, and unclear safety procedures. We built mechanIQs to clear up this confusion and make car maintenance accessible. By empowering people to repair rather than replace, we are actively extending vehicle lifecycles and reducing physical automotive waste, which directly aligns with UVA's sustainability goals to drastically shrink waste footprints.

What it does

Vehicle-Specific Anchoring: Uses the official NHTSA database to tailor every feature to your exact year, make, model, and trim.
Real-Time Parts Catalog: Bypasses stale data using a custom Apify web scraper to pull live pricing and availability directly from RockAuto.
Guided Repairs: Provides step-by-step instructions, complete tool checklists, and critical safety warnings for DIY tasks to prevent botched jobs and wasted physical materials.
Context-Aware AI Assistant: Offers real-time troubleshooting and symptom diagnosis powered by Claude 3.5 Sonnet, which always knows exactly what car you have and what step of the repair you are on.

How we built it

We built the frontend using React 18 and Vite, managing global state so the user's vehicle context persists across the entire application. The backend logic relies on a custom Vite proxy server to securely manage API calls. To align directly with the IT Sustainability track, we embraced the core principles championed by UVA's Green IT Working Group by focusing heavily on computational and energy efficiency. Our custom headless browser scraper on the Apify platform uses intelligent result collection, early job termination, and strict local caching to minimize server load, reduce unnecessary API calls, and lower the overall environmental impact of our computing.

Challenges we ran into

Dynamic Web Scraping: RockAuto relies heavily on JavaScript rendering and anti-bot measures, forcing us to build a custom headless browser Actor on Apify rather than using HTTP requests.
AI Context Management: We had to engineer a sophisticated prompting system to ensure Claude always factored in the user's specific vehicle and current repair step before giving advice.
Data Reliability vs. IT Sustainability: Balancing the speed of our live pricing scraper with accuracy while strictly limiting server timeouts to prevent unnecessary computational energy waste required some optimization. It also benefits the user to not have to wait 30 seconds for a massive drop menu to load.
Secure Architecture: We had to route all third-party services through a local proxy to ensure sensitive API keys were never exposed to the client browser.

Accomplishments that we're proud of

We are proud of engineering a production-grade web scraper that reliably pulls live pricing data while remaining computationally efficient. By keeping our codebase highly optimized and relying on local storage caching, we successfully orchestrated a complex ecosystem of three distinct APIs without generating massive server overhead. Most importantly, our context-aware AI integration feels genuinely helpful rather than generic because it actually understands the exact car and the exact step the user is stuck on, preventing the user from buying the wrong parts and creating more physical and temporal waste.

What we learned

Green IT is Practical: Building our own data pipeline with strict timeout management and caching proved that reducing computational energy consumption actually makes the app faster and cheaper to run.
Rich Context Beats Complex Prompting: Giving the AI exact vehicle specifications and repair steps yielded much better results than generic automotive queries.
Good Design Hides Complexity: Orchestrating React, Apify, Claude, and government databases is technically heavy, but the user only sees the simple interface (great example of abstraction)

What's next for MechanIQs

Green-Certified Mechanic Network: We want to build an ecosystem that bridges the gap between DIYers and professionals. When a repair is too complex or dangerous, mechanIQs will seamlessly route users to local shops that follow strict environmental compliance for disposing of hazardous materials like lithium batteries, motor oil, and coolant.
Circular Economy Parts Scraper: We plan to expand our custom Apify web scraper to pull inventory directly from local salvage yards and refurbished parts suppliers, giving users a cheaper alternative to buying new while completely eliminating the manufacturing carbon footprint of their repair.
Predictive Failure Modeling: By analyzing our users' maintenance logs, we aim to train machine learning models that predict component failures before they cause catastrophic engine damage, keeping the entire vehicle out of the scrapyard.
Automated Recycling Routing: We want to add a geolocation feature that immediately directs users to the nearest specialized recycling center for the exact part they just replaced, ensuring heavy metals and complex plastics never end up in a standard landfill.
Carbon and Waste Impact Tracking: We plan to build a dashboard that calculates the environmental impact of a user's DIY repairs, showing them exactly how much physical waste and carbon emissions they saved by repairing their car instead of buying a new one.