We are focused on developing end-to-end data processing workflows to support the full lifecycle of data. This workflow includes field data acquisition, barcoding, database management, and map visualization. We have been developing technologies to support agro-ecological analyses on a variety of different spatio-temporal scales. We can generate digital maps by collecting crowdsourced assessments of recent satellite imagery backed by expert verification, and then building Machine Learning models to extrapolate predictions throughout the region.
Maps are crucial for many international aid and business projects with national scope, such as agricultural intensification, transport infrastructure, medical surveillance, vaccine delivery, and disaster relief. These maps enable field logistics to be carried out more efficiently, informing routing algorithms with geospatial data. Building footprints may also be used in lieu of government census data that may be non-existent or unreliable.
Project implementers are often directly undertaking this challenge, executing surveys in-person with large crews of local staff, which is extremely time-consuming and expensive. Some national ministries of agriculture have also launched in-house efforts to manually label satellite imagery, but these efforts are often too slow to reach fruition in time for planned interventions. To address these needs, Save the Future has capacity to built automated methods for mapping buildings, transportation infrastructure, water supply network and its associated infrastructure, social services and croplands across nations. By coupling deep learning with crowdsourced collection of precise training data, we can well-approximate building footprints produced by traditional manual methods. Our techniques have been executed for both public and private sector clientele and used for industrial purposes.
Data Collection
Crowdsourced surveying of aerial imagery (satellites) to efficiently generate land cover maps and determine regions of interest such as cropland masks
Randomized survey design: multi-stage sampling algorithms for balancing broad coverage with collection efficiency
Mobile data collection apps for field data entry and real-time data collection.
Sampling protocols tested in Tanzania
