Consensus in Mapping

In this piece, I will specifically talk about community-led field mapping and how terrestrial imagery and consensus can be used to improve the accuracy of data.  

I explain mapping as a method to determine where things are located in space and time. For example, a house is located somewhere in space and time and so is a conversation between two people. Both can be documented or mapped in a similar way.  

Traditionally, we rely on professionals to record places and events and present them to us in various ways, usually through data visualizations or maps. Professionals have the expertise, use highly technical tools in their craft, and abide by rules and standards. Their reputation and work license is on the line. 

It goes without saying that having access to accurate information is essential not only in the realm of geospatial sciences but life in general. Having accurate data enables us to make the right decision at the right time, it allows us to understand the world and its complexities better. 

In recent years, community mapping initiatives exploded worldwide. The cost of technology to collect and display data has reduced significantly and so has the skill necessary to use this technology. Hundreds of community-led initiatives are underway each year where oftentimes inexperienced data collectors produce much-needed information for all sorts of purposes. 

These initiatives are essential to fill in the gap of missing data, especially in places that are not well documented. They take it upon themselves to supply accurate and relevant information or knowledge about a place. 

But how do we ensure that the data they collect is accurate? The repercussions for misrepresenting a place are much lower for a community mapper than the professional surveyor if there are any at all. Leaving out a location or two or even fabricating information isn’t necessarily penalized. 

Community-led field mapping relies on the trust of individuals to do a good enough job. This is because verifying field data is difficult. By difficult, I mean repeating an exercise by walking to every place and ensuring that both location and attribute information is correct is time-consuming. The amount of information that community mapping initiatives produce is often too big for this type of on-the-ground review, and revisiting sites over and over again is too expensive.

Furthermore, these initiatives rely heavily on resource mobilization and funding usually only covers one data collection cycle. When the funding runs out, so does community mapping in most instances. The areas that were underserved before become underserved again and the data can quickly become unreliable. 

We thought a lot about how we can improve field data collection. We asked ourselves: Is there a better way to utilize the power of the community to collect data and keep the information more current, less resource-dependent, more accurate, and easier to update and verify?  

A solution we tested that seems to work well is building a synergy between terrestrial imagery and building consensus through micro-tasking. 

Terrestrial imagery refers to photographs or videos of the Earth’s surface taken from the ground or from low-altitude aircraft. These images can be captured using a variety of techniques and technologies. 

As cartographers, we are particularly interested in the location where the image was captured and the information depicted in it, as the imagery can provide a wealth of useful data.

The most efficient method for obtaining that data from those images is through micro-tasking. Micro-tasking refers to the process of breaking down a larger task into smaller, manageable tasks that can be completed by individuals or a group of people, usually via the internet. A micro-tasking campaign involves distributing terrestrial imagery to a large number of individuals for the purpose of labeling those images. In this way, thousands of images, and their locations, can be reviewed daily. 

The individuals involved in micro-tasking build consensus on every image about what is depicted in an image. In the context of terrestrial imagery, consensus can be understood as the methodology of achieving a mutual agreement on the information depicted in an image and its appropriate representation.

This methodology allows us to scan each area for new information monthly if needed at a relatively low cost, making the data more accurate. The implementation can be carried out by a few people who are needed to manage the cameras and walk the areas of interest, while the bulk of the work is done by the individuals from the comfort of their homes. 

The main benefits of using terrestrial imagery and micro-tasking are the speed with which it can be rolled out, the cost associated with it, and the involvement of a larger community to decide what data is depicted in an image – and by the association at a location -, and how it should be presented. 

In summary, field data collection is essential, especially in poorly represented places. However, fieldwork is expensive because it requires a lot of manpower to carry out, and the cost is oftentimes prohibitive for additional on-the-ground field data validation. Terrestrial imagery and micro-tasking can significantly contribute to the speed with which the data is collected. At the same time, micro-tasking enables building consensus through the engagement of people who collectively decide what is represented at each location, making the data more accurate. 

The world is rapidly changing and the need for the most current and accurate information is increasing. It goes without saying that the methods we use to document this rapid change must follow the trend and evolve or they will falter.