Policy, Ethics and Future Directions in Alleviating Poverty from Space

Traditional poverty data in LMICs rely on infrequent surveys, leaving gaps for timely decisions. Space technologies such as Earth observation and mobile data provide scalable, high-frequency proxies to produce granular poverty maps between censuses. Features like night-time lights, roads, buildings, and agriculture correlate with welfare, and when fused with surveys through “data sandwich” methods, improve precision and coverage. Machine learning, especially CNNs, boosts predictive accuracy and ensures consistent estimates across scales. Policy, Ethics and Future Directions in Alleviating Poverty from Space highlights alignment of EO with SDG mandates, robust data governance, and collaboration across sectors. Ethical use requires privacy-by-design, fairness audits, transparency, and community engagement. Evidence from Africa, Mexico, Bangladesh, Thailand, and Togo shows feasibility, accuracy, and integration into official statistics. Future directions stress multi-temporal monitoring, multimodal fusion, fairness-aware design, and EO-MLOps pipelines, as discussed in this chapter.