{"id":1772,"date":"2015-03-21T06:33:59","date_gmt":"2015-03-21T06:33:59","guid":{"rendered":"http:\/\/www.kam.k.leang.com\/academics\/?page_id=1772"},"modified":"2024-09-12T19:19:18","modified_gmt":"2024-09-12T19:19:18","slug":"uav_pubs","status":"publish","type":"page","link":"http:\/\/www.kam.k.leang.com\/academics\/publications\/uav_pubs\/","title":{"rendered":"Aerial robots and UAV related work"},"content":{"rendered":"
<\/a><\/form>

2026<\/h3>
53.<\/div>
\"Mobile-robotic<\/div>

M. N. Goodell, J. M. Anderson, K. K. Leang\r\n<\/p>

Mobile-robotic Sensor for Estimation and Localization of Multiple Chemical-gas Leaks: A Find- and-Consume Infotaxis Approach<\/a> Journal Article<\/span> <\/p>

In: <\/span>Robotics and Autonomous Systems, <\/span>vol. 202, <\/span>pp. 105431, <\/span>2026<\/span>.<\/p>

Abstract<\/a><\/span> | Links<\/a><\/span> | BibTeX<\/a><\/span><\/p>

@article{GoodellMN_2026,
\r\ntitle = {Mobile-robotic Sensor for Estimation and Localization of Multiple Chemical-gas Leaks: A Find- and-Consume Infotaxis Approach},
\r\nauthor = {M. N. Goodell, J. M. Anderson, K. K. Leang
\r\n},
\r\nurl = {http:\/\/www.kam.k.leang.com\/academics\/pubs\/GoodellMN_2026.pdf},
\r\ndoi = {https:\/\/doi.org\/10.1016\/j.robot.2026.105431},
\r\nyear  = {2026},
\r\ndate = {2026-04-11},
\r\nurldate = {2026-04-11},
\r\njournal = {Robotics and Autonomous Systems},
\r\nvolume = {202},
\r\npages = {105431},
\r\nabstract = {Rapid assessment and localization of accidental or malicious chemical-gas leaks can save lives and minimize environmental impact. An approach is described that quickly estimates and localizes multiple gas leaks using a network of mobile (ground and aerial) robotic sensors. Using the concepts of foraging and consuming food, Bayesian estimation, and information-theoretic motion planning, multiple chemical-gas leaks are found, one source after another. The find-and-consume infotaxis approach makes no assumptions about the total number of sources in a prescribed search area, but it assumes multiple, spatially-distributed gas leaks of the same chemical. Through detailed simulations, metrics such as the correct number of sources identified, the speed of identification, and the source-term estimation accuracy are quantified. These measures are compared to two standard source-finding strategies: (1) raster-scanning and (2) biased-random walk. The results show that the proposed infotaxis method outperforms the two standard approaches, specifically being able to correctly identify up to 10 individual sources 74% of the time on average with an average localization error of approximately 1.2%. Finally, the results from physical experiments using up to four mobile robot platforms equipped with gas sensors (ground and aerial platforms) show successful estimation and localization of three live methane gas leaks, with an average localization error of 4.4%.},
\r\nkeywords = {},
\r\npubstate = {published},
\r\ntppubtype = {article}
\r\n}
\r\n<\/pre><\/div>
Close<\/a><\/p><\/div>
Rapid assessment and localization of accidental or malicious chemical-gas leaks can save lives and minimize environmental impact. An approach is described that quickly estimates and localizes multiple gas leaks using a network of mobile (ground and aerial) robotic sensors. Using the concepts of foraging and consuming food, Bayesian estimation, and information-theoretic motion planning, multiple chemical-gas leaks are found, one source after another. The find-and-consume infotaxis approach makes no assumptions about the total number of sources in a prescribed search area, but it assumes multiple, spatially-distributed gas leaks of the same chemical. Through detailed simulations, metrics such as the correct number of sources identified, the speed of identification, and the source-term estimation accuracy are quantified. These measures are compared to two standard source-finding strategies: (1) raster-scanning and (2) biased-random walk. The results show that the proposed infotaxis method outperforms the two standard approaches, specifically being able to correctly identify up to 10 individual sources 74% of the time on average with an average localization error of approximately 1.2%. Finally, the results from physical experiments using up to four mobile robot platforms equipped with gas sensors (ground and aerial platforms) show successful estimation and localization of three live methane gas leaks, with an average localization error of 4.4%.<\/div>
Close<\/a><\/p><\/div>