Data Collection for Crowdsourced Analysis
Geosurvey Collect allows users to collect data in the field, such as photos and surveys, and easily scale the processing of this data via the crowd. It is the mobile app companion to our Geosurvey crowdsourcing platform. Example applications include:
Our system allows admins to easily launch complex questionnaires where every question can be answered either on mobile or on the web, depending on user needs.
By leveraging web and mobile technology, we hope to bring citizen scientists together and accelerate the collection of geospatial data through worldwide collaboration.
Geosurvey Collect helps projects scale by connecting multiple experts that may be difficult to migrate with field operators that need remote assistance. Below is an example application for diagnosing plant nutrient deficiency and disease. Farm extension agents in Kenya and Tanzania took pictures of ailing crops in their fields, using their mobile phone. When Wi-Fi returns, these pictures are auto-uploaded to the Geosurvey platform for scalable diagnosis by crop health experts. The questions shown on the right-hand panel are interactive and employ conditional logic based on prior answers, walking users through each step of the diagnosis with visual aids along the way.
For our independent analyses of claims of fully-automated diagnosis from other technology companies, please see our study here: PEAT.ai analysis.
In current events at the time of this writing, the fall army worm is devastating maize crops across Sub-Saharan Africa. This started with outbreaks in Zambia during December 2016, and spread through South Africa, Namibia, Kenya and beyond. Estimated yield losses in Kenya alone so far are on the order of $2M USD. Below are screenshots demonstrating the usage of GSC in providing visual and geospatial information about fall army outbreaks to agricultural cooperatives in Kenya that we are partnering with.
Data collected with Geosurvey Collect is only accepted when the mobile device’s GPS variance falls within an acceptable tolerance. Maps are generated as data streams in, providing administrators with visualizations of what in-field users are seeing and where they are seeing it.
When datasets become sufficiently large, machine learning and remote sensing can be used to generate predictive maps. We are very interested in collaboration with passionate citizen scientists and organizations with well-coordinated ground operations.
Below are some directions on how to get started. Please note that Geosurvey Collect is still under active development.
Note that different projects can control (1) who submits to their survey, and (2) who reviews the submitted content. So there are more ongoing projects than an anonymous user can see.
If you think Geosurvey Collect could be useful for your geospatial data collection projects, please contact us to discuss how to best use this technology. We would be happy to hear from you.