High-Level Project Summary
The OceanClean app will help track litter on coastal shores, using artificial intelligence techniques that include data from satellites and applications. I think with the tip you already identified our challenge? Yes, it is the “LEVERAGING AI/ML FOR PLASTIC MARINE DEBRIS”. We propose a mobile app, in which users -- called trackers -- can monitor the concentration of garbage near their location. The tracker will be able to join a team through an event automatically generated by the application, for beach cleaning.When identifying and capturing trash on the beach, trackers will have the option to log into the Debris Tracker app.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
Due to the high concentration of marine plastic waste and waste, monitoring and tracking is of paramount importance, in order to seek and implement solutions for that waste, directing it to a correct location. Therefore, we developed OceanClean, a mobile application to predict the concentration of plastic waste on coastlines and notify the user to a specific location. In this way, the user will be able to report and collect this waste, generating a scoring system, in future plans we can convert these points into discounts and rewards for those who collaborate.
The drebi distribution is obtained through an artificial intelligence method based on the surface roughness of the water (MMSobs) in that region of the globe. Roughness data are obtained from the Cyclone Global Navigation Satellite System (CYGNSS) satellite.
We used as a basis for prediction the work of Madeline C. Evans and Christopher S. Ruf, who developed a method to detect the concentration of drebi in the ocean, through the detection of an anomaly between the roughness-wind relationship. The MSSanom anomaly model involves the calculation of the modeled roughness (MMSmod) for the same location.
The modeled roughness (MMSmod) depends on the wind speed (U) which was calculated using the speed components taken from the NOAA Global Data Assimilation System (GDAS) records. Negative anomaly values indicate the presence of marine litter and the lower this value, the higher the concentration. The point values of the anomaly, são utilizados pelo modelo de inteligência artificial para identificar as regiões costeiras com maior concentração de lixo. O modelo também considera dados abertos do Debris Tracker, como a localização e o tipo de resíduo. No aplicativo, os resultados são mostrados como manchas na costa, destacando áreas com maior volume de resíduo, e que são atualizadas constantemente. Ao clicar nestas manchas, o tracker obtém informações gráficas da distribuição por tipo de resíduo. Com base nestas informações, o sistema do aplicativo cria automaticamente um grupo de mutirão de limpeza, em que os trackers podem se inscrever.
Durante a coleta, os trackers registram as capturas no OceanClean, e os dados são direcionados para o Debris Tracker. Cada tracker possui uma conta, na qual consegue monitorar os seus pontos acumulados de capturas. A partir de parcerias se prevê o desenvolvimento de uma ranking de gamificação envolvendo diferentes tipos de campanhas, como maior quantidade de resíduos plásticos destinados à reciclagem, maior área limpa durante um mutirão ou lixos mais estranhos já coletados na praia.
With the use of OceanClean there will be advantages in monitoring and locating plastic waste and other materials, which will allow the implementation of public measures and their management. In addition, there will be an artificial intelligence, which will forecast the places with the greatest concentration of garbage, providing more time for local officials to get ahead in collection operations and even social activities such as joint efforts by NGOs and the like. Thus, it will be possible to obtain a more strategic and intelligent analysis of coastal waste collection.
Dessa forma, esperamos reduzir a quantidade de lixo nas praias e oceanos, coletando dados para dar insights de onde vem o lixo e como podemos reverter o problema. Além de promover uma maior conscientização da população global em relação à natureza, em especial o oceano e a vida marinha.
We used Python for processing the database and code de artificial intelligence, and ionic for development for the mobile app.
Space Agency Data
The OcenClean artificial intelligence model uses the information from the roughness anomaly (MSSanom) proposed by Madeline C. Evans and Christopher S. Ruf. To calculate the parameter (MSSanom) we need sea roughness data, which were obtained from the Cyclone Global Navigation Satellite System (CYGNSS) Level 2, Version 2.1 with a spatial resolution of 25 km. The system has eight active satellites that provide information on a global level.
The anomaly parameter (MSSanom) also includes wind speed information that was established based on the wind components (u, v) provided by the NOAA Global Data Assimilation System (GDAS) records. The GDAS data are at a spatial resolution of 0.25° and with a frequency of 6 h. We use the same datasets from the study [2] to be able to apply the equations they proposed.
OcenClean's artificial intelligence is also connected to the Debris Tracker open dataset. From this connection, the intelligence model can estimate the distribution by type of residue. On the Debris Tracker platform, collection information is available, such as the location and type of material (plastic, metal, paper,...). This allows refining the classification of the study proposal, as the latter estimates the waste concentration in a generalized way, without specifying the type of waste.
Hackathon Journey
The team was surprised by the challenges in recovering marine waste and how to generate solutions to such a complex problem.
Our team relied on the collaboration of people with multiple different skills, being a diverse team that seeks harmony in all fields, with the objective of generating a better solution. Providing an environment to train soft-skills, important for team problem solving.
One of the biggest inspirations that led us to choose this challenge was the issue of plastic waste, which has been impacting our beaches and oceans. Due to most of the oxygen generated by phytoplankton, pollution negatively influences this O2 production, since with a high amount of plastic waste in the oceans, its temperature tends to rise due to the absorption of heat caused by them. Thus, the change in ocean temperatures causes several problems across the planet, such as: death of marine species, alteration of ocean currents and climate change, in addition to causing serious damage to human health, caused by microplastics.
Therefore, the search for social and environmental solutions is extremely important, as it is useless to develop technologies to terraform Mars and other planets if we do not take care of and solve the problems that exist here on Earth.
References
Data:
CYGNSS: https://doi.org/10.5067/CYGNS-L2X21
GDAS: https://www.ncei.noaa.gov/access/metadata/landing-page/bin/iso?id=gov.noaa.ncdc:C00379
Debristracker: https://debristracker.org/resources
Resources:
PyGNSS: https://github.com/nasa/PyGNSS
[1] Ruf, C., Unwin, M., Dickinson, J., Rose, R., Rose, D., Vincent, M., & Lyons, A. (2013). CYGNSS: Enabling the future of hurricane prediction [remote sensing satellites]. IEEE Geoscience and Remote Sensing Magazine, 1(2), 52-67.
[2] Evans, M. C., & Ruf, C. S. (2021). Toward the Detection and Imaging of Ocean Microplastics With a Spaceborne Radar. IEEE Transactions on Geoscience and Remote Sensing.
[3] Stanley, M. Debris Tracker: Marine. National Oceanic and Atmospheric Administration NOAA, 2010.
Tags
#artificial intelligence, #marine garbage, #app, #tracking
Global Judging
This project has been submitted for consideration during the Judging process.