Inside Unmanned Systems

APR-MAY 2018

Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.

Issue link: https://insideunmanned.epubxp.com/i/969777

Contents of this Issue

Navigation

Page 57 of 67

AIR SOFTWARE 58 April/May 2018 unmanned systems inside C ompanies such a s A mazon and Alphabet hope to use drones to f ly packages to customers. However, programming drones to f ly through cluttered spaces like cities is extraordinarily challeng- ing, even putting aside regulatory concerns. Now, to f ly drones through forests and down streets, researchers are exploring novel tech- niques such as embracing unpredictability or employing artificial intelligence software— work that could help make drones safer to f ly around people. ZIPPING THROUGH FORESTS Delivery services would naturally want drones to f ly as fast as possible to their destinations. However, avoiding obstacles while traveling at high speeds is a computationally complex task, especially for small drones that are limited in how much computing capacity they can carry onboard for real-time analysis. Now a team from MIT's Computer Science and Artif icial Intelligence Laboratory has developed a system called NanoMap that en- ables drones to consistently f ly 20 miles per hour through dense environments like forests and warehouses. The researchers said the secret behind their success was to embrace uncertainty. Many existing strategies aimed at helping drones navigate cluttered environments rely on intricate maps to tell the unmanned air- craft where precisely it is compared to poten- tial obstacles so it can safely move from one point to another. However, if the estimated locations of objects is off by just a slight mar- Fast, Collision Free Urban Flights Photos courtesy of MIT CSAIL, photo by Jonathan How, MIT. To fl y drones through forests and down cluttered city streets, researchers are exploring novel techniques such as embracing unpredictability or employing artifi cial intelligence software. by Charles Q. Choi A drone maneuvers while being trailed by researchers.

Articles in this issue

Archives of this issue

view archives of Inside Unmanned Systems - APR-MAY 2018