RoofShift AI: We Built a Smarter Way to Plan Rooftop Solar Investments in India
Discover how Team DRX developed RoofShift AI, combining India's solar resource maps, weather-aware energy estimates, multi-building budget optimization, performance monitoring, and environmental analytics into one connected renewable-energy planning platform.
Computer Science Student | AI/ML & Full-Stack Developer | AWS & Cloud Computing Enthusiast
RoofShift AI: From Solar Forecasting to Smarter Rooftop Investments in India
An engineering journey by Team DRX for Environmental Hacks 2026, part of the WeMakeDevs Bharat Builds Tour.
What If the Best Solar Investment Isn't the Biggest One?
Imagine a university campus with five buildings.
Every building has a rooftop. Every rooftop receives sunlight. And the university wants to reduce its electricity bills through solar energy.
But there is one problem.
The university has only ₹10 lakh to invest.
Should it install solar panels on the largest building? The rooftop with the highest solar potential? Or the building that consumes the most electricity during daylight hours?
The answer isn't always obvious.
Solar adoption is not just about identifying places with sunlight. It also requires understanding energy generation, consumption patterns, installation costs, and the best possible allocation of limited resources.
That challenge inspired RoofShift AI.
Meet RoofShift AI
RoofShift AI is a rooftop solar intelligence and investment optimization platform designed for educational institutions, hospitals, commercial buildings, factories, and multi-building campuses.
Our vision is to connect the entire solar decision-making process:
Find → Analyze → Forecast → Optimize → Monitor → Improve
Instead of using separate tools for rooftop assessment, solar generation estimates, financial planning, and maintenance, RoofShift AI brings these capabilities into a connected platform.
Our central question is simple:
How can organizations make better solar investments with the budgets and rooftops they already have?
1. Exploring India's Solar Potential Through an Interactive Map
Solar planning starts with geography.
RoofShift AI features an interactive India-focused map where users can search for cities, addresses, landmarks, and geographic coordinates.
The platform supports three levels of exploration:
National exploration: Understand regional solar resource patterns across India.
City and regional exploration: Inspect environmental conditions and local geographic context.
Rooftop exploration: View available OpenStreetMap building footprints and select candidate rooftops for further analysis.
The platform uses geographic information alongside historical solar resource data, including NASA POWER datasets.
An Important Engineering Challenge
One of the challenges we encountered was understanding geographic precision.
A building may appear clearly on a detailed digital map, but satellite-derived solar data can represent a much larger geographic area.
Zooming in does not automatically make the underlying solar measurements more precise.
This distinction matters because customers need trustworthy preliminary estimates, not a beautiful heatmap that promises accuracy it cannot provide.
We therefore distinguish available building geometry from regional solar resource estimates.
2. Solar Power Forecasting and Weather Intelligence
Solar electricity generation changes with environmental conditions.
Cloud cover, irradiance, temperature, system capacity, and installation geometry all influence how much electricity a photovoltaic system may produce.
RoofShift AI connects solar resource information with engineering-based generation estimates.
Our analysis considers relevant inputs such as:
- Solar irradiance
- Temperature
- Environmental conditions
- Installed solar capacity
- System performance assumptions
- Geographic location
The platform helps users examine expected solar electricity generation and understand the influence of environmental factors.
We distinguish long-term historical solar potential from short-term weather forecasting, because those are different forms of information.
Why This Feature Matters
An organization should not decide on a solar installation based only on rooftop area.
It also needs an estimate of how much useful electricity the system may generate over time.
That estimate becomes one of the inputs to the investment optimization process.
3. Multi-Building Budget Optimization: Our Core Innovation
This is where RoofShift AI moves beyond traditional solar assessment.
Most solar-planning workflows focus on identifying suitable sites and estimating how much power they could generate.
But organizations often need to make decisions across several buildings simultaneously.
Consider our original scenario: five buildings and a ₹10 lakh budget.
Each building has different usable rooftop space, potential solar capacity, installation costs, and estimated generation.
How should that budget be distributed?
We approached the problem using constrained mathematical optimization.
The RoofShift AI Budget Optimizer evaluates feasible combinations of rooftop installations while considering available investment and physical capacity constraints.
The User Workflow
Step 1 — Select Candidate Rooftops
Choose the buildings being considered for installation.
Step 2 — Enter the Investment Budget
Specify the maximum investment amount.
Step 3 — Run the Optimizer
The system evaluates candidate allocations according to the selected objective and engineering constraints.
Step 4 — Understand the Recommendation
Review selected buildings, deferred alternatives, installed capacity, budget utilization, and estimated benefits.
Users can also compare the recommended allocation with a simpler baseline strategy.
Why Optimization Matters
The largest rooftop does not necessarily provide the best financial outcome.
Similarly, maximizing raw solar generation may differ from maximizing the value of electricity consumed onsite.
By bringing cost, generation, capacity, and demand information into the planning process, RoofShift AI aims to make solar investment decisions more explainable and financially practical.
Our innovation is not simply predicting solar output. It is helping users decide what to do with that prediction.
4. Generation vs. Consumption: Making Solar Energy More Useful
Electricity generation and electricity consumption do not always happen at the same time.
A solar installation may produce significant electricity around midday, while a building's electricity demand may peak later.
RoofShift AI provides energy-demand analysis to help users understand this relationship.
The platform supports comparisons between modeled solar generation and available electricity-consumption information.
This helps users explore:
- Potential onsite solar consumption
- Remaining grid electricity demand
- Periods of potential solar surplus
- Opportunities to improve solar utilization
The broader planning workflow also explores flexible-load scheduling.
For example, an organization might evaluate whether a water pump or EV charging activity could operate during periods of stronger solar generation.
Such recommendations must respect operational constraints and use transparent calculations.
The purpose is to improve the usefulness of generated electricity—not simply maximize a generation number.
5. Solar Monitoring: Understanding System Performance
Solar planning should not end after installation.
Existing photovoltaic systems need performance visibility, historical analysis, and maintenance support.
RoofShift AI includes an operations interface that separates three monitoring approaches.
Uploaded Historical Data
Users can import compatible generation records, allowing the system to analyze historical trends and selected performance indicators.
Simulation Mode
A clearly labeled testing environment demonstrates how generation, demand, and electricity flows may interact.
Connected Telemetry
The platform provides a connection-oriented workflow for actual monitoring feeds. Genuine live readings require a configured, compatible data source.
This distinction is essential.
Simulated readings are useful for testing, but they should never be presented as real equipment measurements.
The operations interface also includes an energy-flow visualization connecting photovoltaic generation, the inverter, facility loads, optional storage, and the utility grid.
6. Panel Health Inspection and Maintenance Intelligence
Rooftop solar systems can experience issues such as soiling, shading, visible damage, and electrical or thermal anomalies.
Identifying possible problems early can help organizations plan inspections and maintenance.
RoofShift AI supports panel inspection workflows involving uploaded imagery and temperature-survey data.
The thermal analysis workflow can flag temperature differences that may require further investigation.
However, a temperature anomaly by itself does not prove a specific electrical failure.
For that reason, we treat such findings as inspection indicators, not automatic confirmation of equipment defects.
The platform connects these observations with maintenance-case management.
Users can create work orders, review priorities, track unresolved issues, and record maintenance outcomes.
Our longer-term ML direction includes evaluating trained image-classification models using suitable labeled panel images and independently measured test performance.
The important principle is that a model should only be described as trained and validated when reproducible training and evaluation evidence exists.
7. Financial Analysis and Environmental Impact
A renewable-energy investment must make sense both environmentally and financially.
RoofShift AI estimates financial indicators such as:
- Installation investment
- Expected annual electricity generation
- Modeled financial savings
- Available budget allocation
- Estimated simple payback
The platform also estimates potential environmental benefits using electricity-generation assumptions and documented emissions factors.
These estimates can help institutions compare investment scenarios and understand their possible renewable-energy contribution.
We intentionally distinguish modeled environmental benefits from independently verified carbon reductions.
A trustworthy sustainability tool should explain how its numbers were calculated.
8. Our Technology Stack
RoofShift AI uses a web application architecture that combines geographic analysis, numerical computations, and accessible interfaces.
React, TypeScript and Tailwind CSS
Used for the interactive frontend, project workflows, responsive interface, and data visualization.
Python and FastAPI
Used for backend APIs, project records, solar-related calculations, and optimization workflows.
Leaflet and OpenStreetMap
Used for geographic exploration, building-footprint visualization, and rooftop selection.
NASA POWER
Provides historical solar and meteorological information used in resource analysis.
Constraint-Based Optimization
Used to evaluate feasible multi-building solar investment allocations under specified budget and capacity limits.
Reporting and Operational Workflows
Connect the planning results with downloadable summaries, historical performance analysis, and maintenance records.
9. Building with AWS
Environmental Hacks emphasizes making practical use of AWS technologies alongside open-source development.
For our submission, the AWS component must be demonstrated through an actual running service or qualifying open-source integration.
Verified AWS integration: [Insert the exact AWS service or qualifying AWS open-source tool actually used, the feature it supports, and how it was verified.]
Why we selected it: [Explain the engineering problem the integration solves.]
What we learned: [Describe a real setup challenge, technical decision, or improvement resulting from the integration.]
We believe infrastructure choices should be judged by the value they contribute to the application, not by how many cloud services appear in an architecture diagram.
10. What We Learned While Building RoofShift AI
The most valuable engineering lessons came from connecting systems that are often demonstrated independently.
A Map Is Not a Solar Measurement
Geographic visualization must preserve the limitations of the underlying data.
Predictions Need Assumptions
Electricity-generation estimates depend on weather inputs, installation configuration, losses, and model choices.
Optimization Requires Mathematical Discipline
An attractive recommendation is meaningless if it violates the customer's available budget.
Power Is Not Energy
Kilowatts describe power. Kilowatt-hours describe energy over time. Mixing them can create incorrect dashboards and financial calculations.
Live Data Must Actually Be Live
Uploaded historical records and simulation are valuable features, but they are not substitutes for an authenticated live telemetry connection.
Explainability Is a Product Feature
Organizations need to understand why a building was prioritized, not simply receive an unexplained score.
These lessons encouraged us to focus on consistent calculations, data provenance, and meaningful user journeys.
11. What's Next for RoofShift AI?
We see several opportunities to expand the platform:
Weather Forecasting: Improve short-term PV output prediction using forecast weather inputs and appropriately evaluated generation models.
Computer Vision: Train and validate compatible panel-image classifiers for supported visible defects.
Real Equipment Integration: Expand monitoring through authorized inverter and electricity-meter interfaces.
Rooftop Analysis: Improve usable-area and shading estimates using reliable geometric information.
Energy Optimization: Explore richer demand constraints, tariff structures, and flexible-load scheduling.
Decision Transparency: Continue improving the explanation of optimizer results, uncertainty, and model assumptions.
These improvements would build on the existing solar planning and operations foundation.
Closing Thoughts: Smarter Solar Starts with Better Decisions
Solar adoption is often discussed in terms of panel capacity and electricity generation.
But for organizations with limited resources, the deeper question is how to turn solar potential into practical, justified investment decisions.
That is what we are exploring with RoofShift AI.
A platform where geographic information, energy estimates, mathematical optimization, operational analysis, and environmental reporting work together.
Not simply to recommend more solar panels.
But to help organizations make smarter choices about where, when, and how to invest in solar energy.
RoofShift AI — Find. Predict. Optimize. Monitor. Improve.
Developed by Team DRX.
This is an independent student project. References to third-party technologies and event participation do not imply AWS certification, employment, or endorsement.
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article