Introduction:

Over the past year, I have been working as a backend developer on the Field Mapping and Tasking Manager (FMTM) project for the Humanitarian OpenStreetMap Team (HOT). My role has involved system design, feature development, and bug fixing, all aimed at enhancing the efficiency and usability of field mapping for humanitarian purposes. One of my key contributions has been developing and refining the FMTM Splitter —an algorithm designed to streamline field mapping by dividing map Areas of Interest (AOI) into manageable tasks for mappers.

In this post, I will discuss the workings of the FMTM Splitter, the challenges we’ve overcome, and how OpenStreetMap (OSM) data plays a vital role in making this tool effective for humanitarian efforts.

The FMTM Splitter: Two Key Algorithms

The FMTM Splitter is crucial for breaking down large AOIs into smaller, manageable tasks for field mappers. The goal is to ensure that mappers can efficiently collect and update information without having to cross major obstacles like highways or rivers. This makes the mapping process not only faster but also safer and more convenient for field mappers.

We developed two main algorithms for splitting AOIs:

SplitBySquare:

This method splits an AOI into equal-sized square grids based on predefined dimensions, such as 100 meter. It’s a straightforward approach that’s ideal for areas with uniform or relatively simple features.

SplitByAlgorithm:

This more advanced method is designed to split AOIs based on the number of OSM features present, ensuring that each task contains a balanced number of map features such as buildings. Importantly, the algorithm takes into account highways, railways, and waterways to avoid splitting tasks across major infrastructure, ensuring that field mappers don’t have to navigate large barriers while collecting data. When no major barriers are present, the algorithm uses K-Means clustering to divide the AOI based on the specified number of buildings, with additional steps to simplify the task boundaries for ease of use.

Where Does OpenStreetMap Come In?

So, where does OpenStreetMap (OSM) come into the picture? FMTM’s primary mission is to enhance OSM data with detailed, field-verified metadata. We might already have basic data like building outlines or roadways in OSM, but with FMTM, we can enrich these features with critical details collected by field mappers, such as:

• The materials used in construction
• The number of floors in a building
• Typical usage patterns, such as the number of people using a facility

FMTM not only imports OSM data for updating but also allows the inclusion of custom data that might not yet exist in OSM, providing flexibility for field mapping operations. This is particularly important in rapidly evolving situations like disaster relief, where field-collected data plays a crucial role in decision-making.

Lessons Learned and Future Plans

Working on FMTM has provided invaluable insights into both technical and humanitarian challenges. From system scalability to ensuring ease of use for field mappers, the project has reinforced the importance of building tools that can adapt to real-world complexities.

While the current FMTM Splitter algorithm has greatly enhanced field mapping by efficiently dividing Areas of Interest (AOIs), there are opportunities for further optimization. One potential improvement is to introduce a feature that merges smaller tasks with neighboring areas if the created task falls below the desired threshold.

Proposed Feature: Task Merging

When an area is subdivided into tasks, occasionally a few remaining buildings may be grouped into small tasks of 1-2 buildings (as there are no more adjacent buildings to form a larger group). In this scenario, we propose adding an intelligent task-merging mechanism that:

1. Detects Smaller Tasks: Identify any tasks that are below the desired area or contain fewer buildings than specified.
2. Merges with Neighboring Tasks: Combine the smaller task with an adjacent task, ensuring the resulting task is both manageable and aligned with the defined size and desired building number.
3. Retains Boundaries: Ensure that important boundaries like highways, railways, or waterways remain intact during the merging process to avoid inconvenience for field mappers.

Moving forward, I’m excited to continue improving FMTM by exploring new ways to enhance tasking algorithms and integrating more advanced data validation processes. There’s also potential for incorporating machine learning to assist in task creation and AOI splitting based on real-time data.

ODK Entities and Their Role in FMTM

One of the exciting aspects of FMTM is its integration with Open Data Kit (ODK) surveys. ODK Entities are essentially a representation of OSM features, which can be updated based on detailed surveys conducted in the field. This makes it possible to enrich OSM data with high-quality, field-verified information, which is especially important for features that are challenging to verify remotely (e.g. building materials or usage patterns).

Challenges and Solutions:

One of the main challenges we faced while developing the SplitByAlgorithm approach was ensuring that field mappers wouldn’t have to cross major highways, rivers, or other barriers. Incorporating these features into the algorithm was a complex task but crucial for effective mapping. We also had to ensure that when no such features were present, the algorithm could still divide the AOI efficiently using K-Means clustering, balancing the number of OSM features in each task.

Simplifying task boundaries was another key feature we implemented, making the tasks easier to work with in the field. By keeping task boundaries clear and manageable, we help mappers focus on collecting data rather than dealing with confusing or irregular boundaries.

Why This Matters:

By splitting AOIs into manageable tasks, the FMTM Splitter makes it easier for field mappers to add detailed, field-verified metadata to OSM. Whether updating information on schools, hospitals, or other critical infrastructure, this process helps make OSM a more reliable and informative resource for humanitarian work.

The FMTM project aligns closely with HOT’s mission of using open data to support disaster response, development, and community resilience. The improvements we’re making to the tasking manager will help teams in the field work more efficiently and contribute valuable data to the global OSM community.

Call to Action:

If you’re interested in learning more about FMTM or contributing to the project, feel free to reach out! Whether you’re a mapper, developer, or simply someone interested in the intersection of humanitarian work and open data, there are plenty of ways to get involved.