Proposal Writing (half day AM) Hosts: Bob Ryerson and Des Power

One of the continuing issues confronting people in our field is the preparation of successful Responses to Requests for Proposals (RFPs). Over the years the presenters have developed an approach to preparing proposals that has resulted in continuous funding for fifty employees in one case, and in securing many hundreds of millions of dollars for and from government and international development assistant programs in Canada and elsewhere. This half day workshop will address government tenders, development banks and touch on the Grants and Contributions program. It will outline the factors that one should consider for both RFPs and those seeking funding within or from government or international programs. For government or agency funding the workshop will address how to align a proposal with the government’s priorities and the funding source’s legal mandate. This part of the workshop will also consider the rhythm of the funding cycle. Some of the same elements are also important for those responding to RFPs. The workshop will consider how to respond to both fixed price and open price RFPs, addressing (or not addressing) likely competitors, and organizing the response in a way that leads to success. The workshop will close with a period for discussion related to the government focus. It is anticipated that this workshop would be of interest to entrepreneurs, managers in government and industry, grad students, and junior faculty members.

Combining Deep Learning with Editing Tools for 3D Individual-Tree Analysis (half day PM) Host: Zhouxin Xi

The proliferation of terrestrial and UAV lidar data has created an increasing demand for detailed 3D individual-tree analysis. Despite this need, user-friendly, automated processing tools, particularly those offering a “click-and-go” experience remain limited. Although the AI era has unlocked many possibilities, product-level applications that integrate deep learning into 3D forest processing are still rare.

This workshop offers a hands-on solution by integrating deep learning enabled plugins into the open-source 3D editing tool, CloudCompare. Participants will explore common processing tasks using both terrestrial and UAV forest scan datasets. The workshop is divided into two sessions:

Session One: Introduces the core concepts of deep learning and explains how the plugins were developed and integrated with CloudCompare.

Session Two: Demonstrates the practical application of the TreeAIBox plugin for separating tree and ground layers, classifying stem points, detecting tree locations, delineating 3D tree boundaries, and reconstructing 3D branches. This session also includes a detailed tutorial on the manual refinement and interactive annotation of 3D tree boundaries using CloudCompare.

Participants will acquire a comprehensive skill set covering the entire 3D point cloud processing workflow, from initial data handling to achieving high-quality individual-tree delineation. The resulting outputs will be applicable for advanced tree-level analysis, multi-scale data calibration, or serving as reference datasets for training deep learning models.

This workshop is designed for individuals with a foundational understanding of deep learning concepts and some experience working with 3D data. It is suitable for professionals and enthusiasts from various backgrounds, as no specific expertise in remote sensing, GIS, or ecology is required.

Google Earth Engine (full day) Host: Italo Rodrigues

The workshop on Google Earth Engine (GEE) aims to equip participants with essential knowledge of GEE, a robust cloud-based platform for conducting geospatial analysis. Participants will gain insights into navigating the GEE interface, utilizing Landsat and Sentinel-2 imagery, and executing fundamental image processing techniques. The discussion encompasses areas such as image retrieval, visualization techniques, cloud masking methodologies, and the computation of vegetation indices, including NDVI. The participants will delve into land cover classification, acquiring knowledge of both supervised and unsupervised techniques, while evaluating the accuracy of their results.