S-0856
CMLO — AI-Powered Drone Mapping of Coastal and Marine Litter with a Standardised Cleanliness Index
CMLO converts drone imagery of shorelines into standardised, georeferenced litter density maps, material-category classifications and a quantitative cleanliness index. AI model is trained based on over 2.6 million annotated litter items detect objects down to about 2 cm, with results gridded at 10 x 10 m and aligned to international indicators (UN SDG 14.1.1b, EU MSFD). The platform has been deployed across 23 countries in more than 1,800 survey flights. It replaces subjective visual cleanliness ratings with repeatable, quantified and comparable data, enabling departments to target cleaning resources, verify the effectiveness of clean-up operations and generate monitoring reports automatically.
City Management
Environment
Recreation and Culture
Artificial Intelligence (AI)
Data Analytics
Image Recognition
Mobile Technologies
CMLO converts a manual, subjective five-level visual rating into a quantified, auditable and time-comparable index.
How it improves public services:
(1) objective, photo-verifiable cleanliness ratings for every surveyed shoreline, comparable across sites and across time;
(2) density maps by material category, so clean-up resources are deployed where litter actually accumulates;
(3) before/after quantification of clean-up operations, providing the effectiveness data currently missing;
(4) automated bilingual monitoring reports.
Costs and benefits: the solution is flight-based and requires no permanent installation, no hardware on-site and no system integration required.
A one-kilometre shoreline is surveyed in a single flight of under one hour; outputs are delivered within days. A full proof-of-concept covering multiple sites and several survey cycles is about HK$250,000.
LU Frontier Strategies Limited
+85244029740
7/F, Centre Parc, 11 Sheung Yuet Road, Kowloon Bay, Kowloon, Hong Kong
If any government department would like to conduct PoC trial or technology testing on the I&T solution, please contact Smart LAB.