GridLock

Landing page
Coordination Radar
Leaderboard
Project List
Best Matches projects extracted data(from official website)
Extracted data(from official website)

Inspiration

Every year, utility companies plan infrastructure projects worth millions of dollars. Yet those plans often happen in isolation. A transmission line might be rebuilt by one utility, only for another utility or infrastructure provider to perform similar work nearby shortly afterward. The result is duplicated mobilization costs, repeated road closures, additional permitting effort, contractor inefficiencies, and wasted resources.

The same problem exists beyond power transmission. A road can be resurfaced and then reopened weeks later for underground utility work. Fiber providers, water utilities, and energy companies often operate with limited visibility into one another's plans.

We asked a simple question:

What if infrastructure organizations could discover opportunities to coordinate before construction begins?

That idea became GridLock, a platform that turns fragmented public infrastructure plans into a shared coordination radar.

What It Does

GridLock identifies potential coordination opportunities between planned transmission projects.

Using publicly available utility planning documents from multiple companies, GridLock:

Collects and standardizes project information
Maps available geographic data
Analyzes project proximity using GIS calculations
Identifies projects that may benefit from coordination
Surfaces supporting evidence and source documentation
Provides a visual dashboard for investigation

Rather than replacing human judgment, GridLock helps decision makers quickly identify projects that deserve a closer look.

Our current implementation uses 100% real project data from multiple utilities and provides transparent evidence for every opportunity shown.

How We Built It

We built GridLock with Python data pipelines, Shapely for geometry and GIS analysis, Leaflet and JavaScript for the interactive map, and Google Gemini for plain-language explanations. All of it runs on 100% real utility planning data.

The plans told us what was being built, but not always where. So we became part detective, pulling locations out of the documents and marking each one as verified or estimated. Then we measured how close each pair gets, checked whether their schedules line up, and ranked the best opportunities.

For the top matches, we asked the question a project manager would ask: how much could this actually save?

$$\text{Savings} \approx \text{Shared work} \times \text{Setup cost \%} \times \text{Portion you can avoid}$$

For our #1 pair, that's roughly $60k–$450k for every $10M of shared work. The savings would come from things like one crane trip instead of two, or one staging yard instead of two.

Our workflow consisted of:

Collecting public project planning data
Normalizing inconsistent formats across utilities
Extracting and enriching geographic information
Creating verified and estimated geometry datasets
Calculating project to project proximity using GIS analysis
Ranking potential coordination opportunities
Building an interactive dashboard to explore results
Integrating AI assisted explanations to help users understand findings

A core design principle guided the project:

AI explains. GIS verifies. Humans decide.

Challenges We Ran Into

One of the biggest challenges was working with real world infrastructure data.

Unlike clean academic datasets, public utility planning documents are inconsistent, incomplete, and spread across many sources. Simply extracting the information and building a reliable data pipeline required significant effort.

Other challenges included:

Reconciling data from multiple utilities
Recovering usable geographic information from planning documents
Distinguishing verified geography from estimated geography
Building accurate opportunity scoring without introducing false positives
Ensuring that timing and geographic proximity both aligned
Creating meaningful cost saving estimates that were realistic for the industry
Presenting complex GIS analysis in a way that nontechnical users could immediately understand

Making the platform trustworthy was just as important as making it functional.

Accomplishments That We're Proud Of

We're especially proud that:

The platform uses real utility planning data rather than synthetic examples
Every opportunity is backed by evidence and source documentation
Our GIS analysis is transparent and explainable
We successfully created a working end to end data pipeline
We built an interactive coordination radar and project explorer
We integrated AI without allowing it to fabricate results
We created a system that helps users investigate opportunities rather than blindly trust recommendations

Most importantly, we're proud that GridLock demonstrates how real world infrastructure planning can become more coordinated, efficient, and transparent.

What We Learned

This project taught us that the hardest problems are often not the algorithms themselves, but the data.

We learned:

How difficult it is to standardize infrastructure planning datasets
The importance of provenance and traceability
Practical GIS analysis techniques
How to build explainable systems instead of black box systems
How to incorporate AI responsibly into decision support tools
How much value can be unlocked when disconnected datasets are brought together

We also learned that effective visualization is just as important as accurate analysis. Even the best insights are only useful if people can quickly understand and act on them.

What's Next for GridLock

This is only the beginning.

Future versions of GridLock could:

Expand to additional utilities and infrastructure sectors
Incorporate historical project outcomes
Predict potential delays and coordination risks
Analyze contractor and resource utilization patterns
Identify regions with recurring infrastructure bottlenecks
Use weather, flooding, hurricane, wildfire, and other environmental data to anticipate future infrastructure activity
Provide regional preparedness insights for utilities, government agencies, contractors, and emergency planners
Deliver AI generated executive briefings and planning recommendations

Our vision is a future where infrastructure organizations can coordinate proactively instead of reactively, reducing costs, minimizing disruption, and building more resilient communities.

GridLock transforms fragmented infrastructure plans into actionable coordination intelligence before construction even begins.