Projects

In late 2024, Saiwa Co and Ducks Unlimited Canada launched a project to enhance UAV-based detection and surveillance of the invasive aquatic plant Water Soldier (Stratiotes Aloides) using advanced image processing and deep learning. The team addressed challenges from dense and submerged vegetation, environmental noise, and irregular plant shapes by reformulating the task as semantic segmentation. The project delivered a robust model achieving 95.52% pixel accuracy, enabling accurate monitoring of both isolated plants and dense colonies, supporting early intervention, efficient management, and a scalable foundation for future multispectral enhancements.

In early 2025, Saiwa and Airwyse collaborated to automate counting of pollination bags in pine orchards using autonomous drones and AI-driven image analysis. This innovative solution replaced labor-intensive manual counting, improving accuracy and efficiency in yield prediction. Tested on 44 pine trees, the system demonstrated promising results, paving the way for scalable orchard management automation.

Powered by Sairone and funded by OAFRI, this project leverages AI and drone imagery to detect herbicide-tolerant Canada Fleabane in Ontario soybean fields. Designed to empower farmers with no-code tools, it transforms raw agricultural images into actionable insights—enabling geotagged weed detection, herbicide mapping, and data ownership. By lowering tech adoption barriers, this initiative supports sustainable, privacy-first farming across Canada.

To enhance Ducks Unlimited Canada (DUC)‘s capability for UAV-based surveillance of European Water Chestnut (EWC) through machine learning, we at Saiwa have previously implemented the initial version of the EWC detector software. In the second stage, we are in the process of finalizing the product’s features and upgrading its interfaces.The two primary features to be incorporated in this stage are as follows: Incremental learning for gradually training the deep network over time. This feature enables us to rectify false positive and false negative detections over time. Reporting the 3D universal coordinates of EWC locations using drone configuration and temporal GPS data.