Daffodil Software
Trusted partner for custom software development
Trusted partner for custom software development
Green City Watch is a geospatial AI firm specializing in urban ecological engineering. It leverages the Industrial Internet of Things (IIoT) technology to empower municipalities and local councils with data-driven urban ecology. Green City Watch empowers 30+ (mega)cities, from Boston to Amsterdam to Jakarta, to understand, monitor, and improve their urban forests through data-driven decision-making. Their TreeTect™ technology provides municipalities and urban foresters with a better overview of their tree inventory. By combining Industry 4.0 technologies with satellites, LiDAR, and drones, TreeTect™ allows municipalities to draw actionable insights about green space in near real-time.
The Situation
Urban foresters are increasingly challenged by the complexities of changing urban ecosystems, hindered by low budgets, outdated data, and low-tech solutions. In cities like Amsterdam Boston, where urban trees have a lifespan of only about 13 years compared to over 100 years for rural trees, there is a pressing need to enhance tree longevity. Green City Watch envisioned leveraging geospatial AI—an integration of ecological engineering, machine learning, and remote sensing—to improve urban tree management.
To support urban foresters, Green City Watch proposed developing an AI-based solution that would create a comprehensive tree inventory for Amsterdam and Boston. This inventory would enable foresters to derive actionable insights about green spaces. Utilizing high-resolution geospatial data captured from WorldView-1 and WorldView-3 satellites, they aimed to calculate the Normalized Difference Vegetation Index (NDVI), a key indicator of vegetation health.
Seeking a technology partner to realize this vision, Green City Watch chose Daffodil Software for its proven expertise in AI technologies. The project focused on two main objectives: to design an AI solution capable of identifying green areas and providing a detailed overview from varied sensors, and to accurately identify 80-90% of trees within a defined area with a 3-meter error radius, while also determining their geo-coordinates and crown sizes. This initiative aims to significantly enhance the management and longevity of urban trees.
The project began with Daffodil's business analytics and software architects outlining the solution's architecture, refining functional requirements, and developing a product vision and roadmap. Once finalized, the team selected core technologies such as QGIS, Leaflet, Node.js, AWS, TensorFlow, and GBDX for development.
To locate urban trees, Daffodil employed object detection techniques, utilizing satellite-captured spatial data stored in GeoTIFF files. Given that different tree species are identified across various bands, multiple machine learning models were developed, using ensemble modeling to enhance prediction accuracy. This approach aggregates predictions from diverse models to achieve a final outcome.
By combining three bands, eleven models were created to identify tree species and locations, successfully identifying over 90% of trees in Amsterdam and Boston across various conditions. To ensure scalability, different band combinations were designed for different geographies. Daffodil also utilized various AWS cloud services to optimize resources and costs, including AWS SES, ECR, Sagemaker, S3, and Lambda for efficient resource management.
The client has been extremely satisfied by the way Daffodil has executed their vision and have planned for further updates to the system. The TreeTect technology provides municipalities and urban foresters with a better overview of their tree inventory. By combining Industry 4.0 technologies with satellites, LiDAR, and drones, TreeTect allows municipalities to draw actionable insights about green space in near real-time.
Key stats:
