AeroAstro Annual 8
Annual Report 2010-2011
23 WHERE AM I? Absent External Positioning Systems, UAVs Need to be Clever Learners UAV guidance and control has traditionally depended on both an external positioning signal such as GPS for position information, and an existing map for planning flight paths. This depen- dence has restricted UAVs to applications where the vehicle can operate at altitude, ensuring good GPS signals and few obstacles. OPERATIONAL CHALLENGES INCREASING Operational requirements for UAVs, especially micro aerial vehicles (MAVs), are expanding to include flight at low altitude in complex, populated environments. Search and rescue, surveillance, and other tasks may require MAVs to reliably fly through the “urban canyon” and enter buildings. However, most indoor environments and many parts of urban envi- ronments remain without access to external positioning systems such as GPS, thus limiting autonomous MAVs’ ability to fly through these areas. To overcome the absence of external position systems, onboard sensors have been used successfully for navigation in a number of autonomous vehicle applications. For operation in unknown environments, these sensors allow a robot to explore the environment, collect sensor data, such as camera images or data from a ranging sensor, and use statistical models to assemble the data into a globally consistent map of the world. Once enough data is collected WHERE AM I? Absent external positioning systems, UAVs need to be clever learners By Nicholas Roy and Jonathan P. How Autonomous unmanned aerial vehicles are increasingly a part of military and civilian ﬂight operations. Examples range from surveillance over Iraq and Afghanistan, to critical use in moni- toring the Fukushima nuclear reactor and disaster relief coordination in response to the Haitian earthquake.