惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

云风的 BLOG
云风的 BLOG
M
MIT News - Artificial intelligence
Recent Announcements
Recent Announcements
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Stack Overflow Blog
Stack Overflow Blog
J
Java Code Geeks
Microsoft Azure Blog
Microsoft Azure Blog
罗磊的独立博客
博客园 - 【当耐特】
H
Help Net Security
腾讯CDC
大猫的无限游戏
大猫的无限游戏
GbyAI
GbyAI
Last Week in AI
Last Week in AI
Jina AI
Jina AI
博客园 - 聂微东
Blog — PlanetScale
Blog — PlanetScale
A
About on SuperTechFans
Apple Machine Learning Research
Apple Machine Learning Research
P
Proofpoint News Feed
Y
Y Combinator Blog
C
Check Point Blog
博客园 - 司徒正美
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
We Built a 3-Layer Audit Trail (AI + GPS + Blockchain) to...
Chad Buliamu · 2026-04-29 · via DEV Community

Most environmental charities ask you to trust them. We built a system so you don't have to.

I'm Valentin, co-founder of Thriving Planet Association, a Swiss nonprofit that recovers ocean plastic. But this isn't a fundraising post dressed up as a tech article. It's about a real engineering problem we had to solve: how do you make environmental impact verifiable, tamper-proof, and visible to anyone who funds it?
Here's what we built and why it matters for anyone working at the intersection of technology and social good.

** The Problem: "Trust Us" Doesn't Scale**

The environmental sector has a transparency crisis.
Donors are increasingly skeptical and rightfully so. Terms like "carbon offset" and "plastic neutral" have been so thoroughly abused by corporate greenwashing that many people have stopped believing environmental claims altogether.
For ocean conservation specifically, the problem is acute:

How do you prove that a specific kilogram of plastic was collected, rather than just estimated?
How do you ensure the GPS coordinates of a cleanup aren't fabricated?
How do you give a donor in Zurich verifiable proof of what happened on a beach in Southeast Asia?

We asked ourselves these questions before we ran a single collection. The answer was a 3-layer technical audit trail.

The Solution: AI + GPS + Blockchain

We call it the 3-Layer Audit Trail. Here's how each layer works:

** Layer 1 AI Verification (Weight & Classification)**

Every gram of plastic collected is weighed and photographed at the point of collection. A computer vision model classifies plastic by type (PET, HDPE, LDPE, etc.) and cross-references the weight against the visual volume of material. This eliminates human rounding errors and makes it significantly harder to inflate collection numbers.

Why this matters: Plastic classification data is also valuable for downstream recycling partners — knowing the composition of collected material improves recycling yield and creates a secondary accountability loop.

Layer 2 GPS Timestamping

Each collection event is tagged with:

Precise GPS coordinates (latitude/longitude)
Timestamp (UTC)
Collector ID
Photo evidence

This data is captured at the moment of collection and cannot be retroactively altered. Donors receive a digital certificate showing the exact location where their sponsored kilogram was recovered.
Why this matters: GPS coordinates are independently verifiable. Anyone can open Google Maps and see the collection site. There's nowhere to hide.

** Layer 3 Blockchain Immutability**

The AI classification result, weight, GPS coordinates, and timestamp are hashed and written to a blockchain ledger. Once written, this record cannot be modified, deleted, or backdated.
This gives donors something unprecedented in charitable giving: an immutable, publicly auditable record of their impact.

** What This Looks Like in Practice**

When a donor contributes CHF 100 to our campaign, here's what happens:

  1. Collection team arrives at verified site (GPS tagged)
  2. Plastic is collected and weighed on calibrated scales
  3. The AI model classifies and validates the weight/volume ratio
  4. GPS + weight + timestamp → hashed and written to blockchain
  5. Donor receives digital certificate with:
    • Collection site coordinates
    • Weight recovered
    • Blockchain transaction ID (publicly auditable)
    • Photo evidence from the collection No black box. No "trust us." Just verifiable data.

** Why We're Sharing This on Dev. to**

We're currently running a CHF 1,000 crowdfunding campaign to launch a new verified collection point. But more than the funding, we want to start a conversation in the tech community about what verified impact infrastructure should look like.
We believe the tools exist today — computer vision, GPS, blockchain — to make greenwashing technically impossible. Not harder. Impossible.
If you're a developer, data scientist, or blockchain engineer who's thought about applying your skills to environmental problems, we'd love to connect.
And if you want to support the campaign directly:

CHF 25 = 1kg of plastic recovered and verified, with full audit trail
CHF 100 = 4kg + a digital certificate with exact GPS coordinates of your impact

👉 Support the campaign on GoFundMe

**https://www.gofundme.com/f/thriving-planet-associations?lang=en_GB

Open Questions We're Still Working On

We're not claiming to have solved everything. Here are the hard problems we're still wrestling with:

The Oracle problem with blockchain is only as trustworthy as the data written to it. If a collector falsifies the weight before it's hashed, the chain records a false truth immutably. We're exploring IoT-connected scales as a mitigation.
Scalability writing every collection event to a blockchain has cost and throughput implications. We're evaluating layer-2 solutions and batch hashing approaches.
Accessibility blockchain certificates are meaningless to most donors without a readable interface. We're building a human-friendly dashboard that abstracts the technical complexity.

If you've worked on similar problems in supply chain verification, carbon accounting, or impact measurement, we'd genuinely love to hear from you in the comments.

** Final Thought**

The technology to make environmental impact fully transparent already exists. The question is whether organisations are willing to build the infrastructure to use it even when opacity is more convenient.
We chose transparency. And we're betting that the people who fund conservation work will choose it too.

Thanks for reading.