Chronos

Chronos is a self-reported tool that pairs street photos taken years apart using AI to detect changes in urban environments. The author states it uses computer vision, geospatial logic, and structured outputs from large language models to map durable decay not recorded by cities. It runs locally with Python and JavaScript, and the system is built around idempotent stages and a chunk-and-stitch pipeline for API data fetching.

The project appears to be a proof-of-concept or prototype built in a hackathon setting. The author describes it as a tool that could become an accountability layer for public space, but no evidence of revenue, customers, or adoption is provided. Key commercial questions remain: whether the system can scale beyond a single developer's local instance, how it would integrate with municipal workflows, and whether there are viable monetization paths.

The description states that Chronos:

• Pairs street-level photos taken years apart using geospatial logic and vision models.
• Uses AI to verify what changed between images.
• Maps verified changes on an interactive map with before-and-after sliders and 360° views.
• Operates locally with Python, FastAPI, SQLite, JavaScript, MapLibre GL, and Mapillary JS.
• Runs as a CLI tool with three commands: ingest, inspect, and serve.
• Stores all data in SQLite and is designed to be idempotent.

It is described as a system that uses structured outputs from vision models, with a confidence floor in code to avoid false positives. The system also includes a custom pairing algorithm based on distance, heading, time gap, and sequence ID, and a quadtree-based fetch strategy for Mapillary API data.

Inference Chronos appears to be a local prototype that processes existing imagery from platforms like Mapillary to detect changes in urban infrastructure over time.

The author states:

• Cities track construction and demolition but not small, durable decay.
• The idea is to use existing street-level imagery to build an accountability layer for public space.
• It aims to map “the layer underneath” permit records — subtle changes like faded crosswalks or cracked curb ramps.

Inference Chronos positions itself as a tool that fills a gap in urban data by leveraging existing visual data, rather than collecting new data. The author frames it as an accountability mechanism for public infrastructure, not a commercial product or platform.

The description states:

• The target is cities and municipal departments.
• It aims to provide early warning on maintenance backlogs, broken curb ramps, and faded crossings.
• It could be used by planners and researchers for longitudinal analysis.

Inference The primary customer is likely city governments or public agencies that manage infrastructure. The ICP appears to be urban planners, maintenance departments, and researchers who want to track changes in public space over time.

Not evidenced.

The description does not state anything about pricing, monetization, or a business model beyond the author’s own use case or vision for future expansion. No revenue streams, customer acquisition plans, or commercial partnerships are mentioned.

The author states:

• Built with Python 3.12, FastAPI, SQLite, JavaScript, MapLibre GL, and Mapillary JS.
• Uses structured outputs from LLMs (GPT-4o, GPT-5.6) to ensure reliable judgments.
• Implements a custom pairing algorithm based on geospatial logic.
• Uses a quadtree-based API fetch strategy to handle large urban areas.
• Designed to be idempotent and run locally with no build step.

Inference The system is built for reliability, testability, and scalability in data ingestion. It uses modern tools and constraints (like structured outputs) to ensure trustworthiness of AI judgments.

Not evidenced.

There is no mention of revenue, customers, user adoption, or product usage beyond the author’s own development and demo deployment. The system is described as a prototype built in a hackathon, with no evidence of real-world application or traction.

Not evidenced.

The description does not mention any competitors or existing tools that address similar problems. No market analysis or competitive positioning is provided.

• The system is described as running locally and being a prototype, not a scalable product.
• It relies on Mapillary API access, which may be limited or costly.
• The author’s own write-up notes that the live demo is read-only and does not include real-time features due to API cost concerns.
• No evidence of monetization, customer base, or commercial viability.
• The system is built for a single developer and lacks any indication of team or infrastructure scaling.

Inference Chronos is a proof-of-concept with no demonstrated traction or commercial viability. It may not be ready for production use or integration into city workflows.

1. What is the actual data volume and geographic scope that Chronos can process in its current form?
2. Has there been any testing or feedback from municipal departments or urban planners?
3. How does the system handle edge cases like weather, lighting, or seasonal changes?
4. Are there plans to integrate with existing city data systems or permit databases?
5. What are the real-world costs of running the system at scale (API usage, compute, etc.)?
6. Is there a path to monetization or partnership with cities or public agencies?