{"type":"ScholarlyArticle","title":"Bayesian Active Edge Evaluation on Expensive Graphs","authors":[{"name":"Sanjiban Choudhury","slug":"sanjiban-choudhury","given":"Sanjiban","family":"Choudhury","affiliation":"School of Computer Science and Engineering , University of Washington , USA"},{"name":"Siddhartha Srinivasa","slug":"siddhartha-srinivasa","given":"Siddhartha","family":"Srinivasa","affiliation":"School of Computer Science and Engineering , University of Washington , USA"},{"name":"Sebastian Scherer","slug":"sebastian-scherer","given":"Sebastian","family":"Scherer","affiliation":"The Robotics Institute , Carnegie Mellon University , USA"}],"abstract":"We consider the problem of real-time motion planning that requires evaluating a minimal number of edges on a graph to quickly discover collision-free paths. Evaluating edges is expensive, both for robots with complex geometries like robot arms, and for robots sensing the world online like UAVs. Until now, this challenge has been addressed via laziness, i.e. deferring edge evaluation until absolutely necessary, with the hope that edges turn out to be valid. However, all edges are not alike in value - some have a lot of potentially good paths flowing through them, and some others encode the likelihood of neighbouring edges being valid. This leads to our key insight - instead of passive laziness, we can actively choose edges that reduce the uncertainty about the validity of paths. We show that this is equivalent to the Bayesian active learning paradigm of decision region determination (DRD). However, the DRD problem is not only combinatorially hard but also requires explicit enumeration of all possible worlds. We propose a novel framework that combines two DRD algorithms, DIRECT and BISECT, to overcome both issues. We show that our approach outperforms several state-of-the-art algorithms on a spectrum of planning problems for mobile robots, manipulators and autonomous helicopters.","doi":"10.24963/ijcai.2018/679","url":"https://ijcai.org/proceedings/2018/679","pdf":"https://ijcai.org/proceedings/2018/0679.pdf","markdown":"https://ijcai.org/proceedings/2018/679.md","bibtex":"https://ijcai.org/proceedings/2018/bibtex/679","volume":{"slug":"2018","title":"Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2018","url":"https://ijcai.org/proceedings/2018/","year":2018,"location":"Stockholm","dates":"13-19 July 2018","editors":"Jérôme Lang","isbn":"978-0-9992411-2-7","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2018"},"section":"Robotics","track":"Main track","number":679,"pages":{"first":4890,"last":4897},"keywords":["Planning and Scheduling: Real-time Planning","Robotics: Learning in Robotics","Machine Learning: Active Learning","Planning and Scheduling: Robot Planning","Robotics: Manipulation","Robotics: Motion and Path Planning"],"videos":[],"references":[{"position":1,"raw":"[Bialkowski et al., 2013] Joshua Bialkowski, Michael Otte, and Emilio Frazzoli. Free-configuration biased sampling for motion planning. In IROS, 2013.","title":"Free-configuration biased sampling for motion planning","authors":"Joshua Bialkowski, Michael Otte, Emilio Frazzoli","year":"2013","venue":"IROS","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bialkowski et al., 2016] Joshua Bialkowski, Michael Otte, Sertac Karaman, and Emilio Frazzoli. Efficient collision checking in sampling-based motion planning via safety certificates. The International Journal of Robotics Re- search, 35(7):767-796, 2016.","title":"Efficient collision checking in sampling-based motion planning via safety certificates","authors":"Joshua Bialkowski, Michael Otte, Sertac Karaman, Emilio Frazzoli","year":"2016","venue":"The International Journal of Robotics Research","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Bohlin and Kavraki, 2000] Robert Bohlin and Lydia E Kavraki. Path planning using lazy prm. In ICRA, 2000.","title":"Path planning using lazy prm","authors":"Kavraki Bohlin, Robert Bohlin, Lydia E Kavraki","year":"2000","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Burns and Brock, 2005] Brendan Burns and Oliver Brock. Sampling-based motion planning using predictive models. In ICRA, 2005.","title":"Sampling-based motion planning using predictive models","authors":"Brock ; Burns, Brendan Burns, Oliver Brock","year":"2005","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Chaloner and Verdinelli, 1995] Kathryn Chaloner and Is- abella Verdinelli. Bayesian experimental design: A review. Statistical Science, pages 273-304, 1995.","title":"Kathryn Chaloner and Isabella Verdinelli. Bayesian experimental design: A review","authors":"Verdinelli Chaloner","year":"1995","venue":"Statistical Science","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Chen et al., 2015] Yuxin Chen, Shervin Javdani, Amin Kar- basi, Drew Bagnell, Siddhartha Srinivasa, and Andreas Krause. Submodular surrogates for value of information. In AAAI, 2015.","title":"Submodular surrogates for value of information","authors":"Yuxin Chen, Shervin Javdani, Amin Karbasi, Drew Bagnell, Siddhartha Srinivasa, Andreas Krause","year":"2015","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Choudhury et al., 2014] Sanjiban Choudhury, Sankalp Arora, and Sebastian Scherer. The planner ensemble and trajectory executive: A high performance motion planning system with guaranteed safety. In AHS 70th Annual Forum, 2014.","title":"The planner ensemble and trajectory executive: A high performance motion planning system with guaranteed safety","authors":"Choudhury","year":"2014","venue":"AHS 70th Annual Forum","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Choudhury et al., 2016a] Sanjiban Choudhury, Jonathan D. Gammell, Timothy D. Barfoot, Siddhartha Srinivasa, and Sebastian Scherer. Regionally accelerated batch informed trees (rabit*): A framework to integrate local information into optimal path planning. In ICRA, 2016.","title":"Regionally accelerated batch informed trees (rabit*): A framework to integrate local information into optimal path planning","authors":"Choudhury","year":"2016","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Choudhury et al., 2016b] Shushman Choudhury, Christo- pher M Dellin, and Siddhartha S Srinivasa. Pareto-optimal search over configuration space beliefs for anytime motion planning. In IROS, 2016.","title":"Pareto-optimal search over configuration space beliefs for anytime motion planning","authors":"Choudhury","year":"2016","venue":"IROS","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Choudhury et al., 2017a] Sanjiban Choudhury, Shervin JAvdani, Siddhartha Srinivasa, and Sebastian Scherer. Near-optimal edge evaluation in explicit generalized binomial graphs. In NIPS, 2017.","title":"Near-optimal edge evaluation in explicit generalized binomial graphs","authors":"Choudhury","year":"2017","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Choudhury et al., 2017b] Sanjiban Choudhury, Shervin JAvdani, Siddhartha Srinivasa, and Sebastian Scherer. Near-optimal edge evaluation in explicit generalized binomial graphs. Arxiv, 2017.","title":"Near-optimal edge evaluation in explicit generalized binomial graphs","authors":"Choudhury","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Cohen et al., 2015] Benjamin Cohen, Mike Phillips, and Maxim Likhachev. Planning single-arm manipulations with n-arm robots. In Eigth Annual Symposium on Com- binatorial Search, 2015.","title":"Planning single-arm manipulations with n-arm robots","authors":"Benjamin Cohen, Mike Phillips, Maxim Likhachev","year":"2015","venue":"Eigth Annual Symposium on Combinatorial Search","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Dasgupta, 2004] Sanjoy Dasgupta. Analysis of a greedy ac- tive learning strategy. In NIPS, 2004.","title":"Analysis of a greedy active learning strategy","authors":"Sanjoy Dasgupta","year":"2004","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Dellin and Srinivasa, 2016] Christopher M Dellin and Sid- dhartha S Srinivasa. A unifying formalism for shortest path problems with expensive edge evaluations via lazy best-first search over paths with edge selectors. In ICAPS, 2016.","title":"A unifying formalism for shortest path problems with expensive edge evaluations via lazy best-first search over paths with edge selectors","authors":"M Christopher, Siddhartha S Dellin, Srinivasa","year":"2016","venue":"ICAPS","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Dellin et al., 2016] Christopher M Dellin, Kyle Strabala, G Clark Haynes, David Stager, and Siddhartha S Srini- vasa. Guided manipulation planning at the darpa robotics challenge trials. In Experimental Robotics, 2016.","title":"Guided manipulation planning at the darpa robotics challenge trials","authors":"Kyle Christopher M Dellin, Clark Strabala, David Haynes, Siddhartha S Stager, Srinivasa","year":"2016","venue":"Experimental Robotics","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Gammell et al., 2015] Jonathan D. Gammell, Siddhartha S. Srinivasa, and Timothy D. Barfoot. Batch Informed Trees: Sampling-based optimal planning via heuristically guided search of random geometric graphs. In ICRA, 2015.","title":"Batch Informed Trees: Sampling-based optimal planning via heuristically guided search of random geometric graphs","authors":"Jonathan D Gammell, Siddhartha S Srinivasa, Timothy D Barfoot","year":"2015","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Golovin and Krause, 2011] Daniel Golovin and Andreas Krause. Adaptive submodularity: Theory and applications in active learning and stochastic optimization. Journal of Artificial Intelligence Research, 2011.","title":"Adaptive submodularity: Theory and applications in active learning and stochastic optimization","authors":"Golovin, Krause, Daniel Golovin, Andreas Krause","year":"2011","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Golovin et al., 2010] Daniel Golovin, Andreas Krause, and Debajyoti Ray. Near-optimal bayesian active learning with noisy observations. In NIPS, 2010.","title":"Near-optimal bayesian active learning with noisy observations","authors":"Andreas Daniel Golovin, Debajyoti Krause, Ray","year":"2010","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Hauser, 2015] Kris Hauser. Lazy collision checking in asymptotically-optimal motion planning. In ICRA, 2015.","title":"Lazy collision checking in asymptotically-optimal motion planning","authors":"Kris Hauser","year":"2015","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Howard, 1966] Ronald A Howard. Information value the- ory. IEEE Tran. Systems Science Cybernetics, 1966.","title":"Information value theory","authors":"A Ronald, Howard","year":"1966","venue":"IEEE Tran. Systems Science Cybernetics","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Hsu et al., 1997] David Hsu, J-C Latombe, and Rajeev Mot- wani. Path planning in expansive configuration spaces. In ICRA, 1997.","title":"Path planning in expansive configuration spaces","authors":"David Hsu, J-C Latombe, Rajeev Motwani","year":"1997","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Huh and Lee, 2016] Jinwook Huh and Daniel D Lee. Learn- ing high-dimensional mixture models for fast collision de- tection in rapidly-exploring random trees. In ICRA, 2016.","title":"Learning high-dimensional mixture models for fast collision detection in rapidly-exploring random trees","authors":"Lee ; Huh, Jinwook Huh, Daniel D Lee","year":"2016","venue":"ICRA","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Javdani et al., 2014] Shervin Javdani, Yuxin Chen, Amin Karbasi, Andreas Krause, Drew Bagnell, and Siddhartha Srinivasa. Near optimal bayesian active learning for deci- sion making. In AISTATS, 2014.","title":"Near optimal bayesian active learning for decision making","authors":"Shervin Javdani, Yuxin Chen, Amin Karbasi, Andreas Krause, Drew Bagnell, Siddhartha Srinivasa","year":"2014","venue":"AISTATS","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Kavraki et al., 1996] L.E. Kavraki, P. Svestka, J.C. Latombe, and M.H. Overmars. Probabilistic roadmaps for path planning in high-dimensional configuration spaces. Robotics and Automation, IEEE Transactions on, 12(4):566-580, 1996.","title":"Probabilistic roadmaps for path planning in high-dimensional configuration spaces","authors":"L E Kavraki, P Svestka, J C Latombe, M H Overmars","year":"1996","venue":"Robotics and Automation","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Kononenko, 2001] Igor Kononenko. Machine learning for medical diagnosis: History, state of the art and perspective. Artificial Intelligence in Medicine, 2001.","title":"Machine learning for medical diagnosis: History, state of the art and perspective","authors":"Igor Kononenko, Kononenko","year":"2001","venue":"Artificial Intelligence in Medicine","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Krause and Guestrin, 2009] Andreas Krause and Carlos Guestrin. Optimal value of information in graphical mod- els. Journal of Artificial Intelligence Research, 35:557- 591, 2009.","title":"Optimal value of information in graphical models","authors":"Guestrin Krause, Andreas Krause, Carlos Guestrin","year":"2009","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Lacevic et al., 2016] Bakir Lacevic, Dinko Osmankovic, and Adnan Ademovic. Burs of free c-space: a novel struc- ture for path planning. In ICRA. IEEE, 2016.","title":"Burs of free c-space: a novel structure for path planning","authors":"Bakir Lacevic, Dinko Osmankovic, Adnan Ademovic","year":"2016","venue":"ICRA. IEEE","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Narayanan and Likhachev, 2017] Venkatraman Narayanan and Maxim Likhachev. Heuristic search on graphs with ex- istence priors for expensive-to-evaluate edges. In ICAPS, 2017.","title":"Heuristic search on graphs with existence priors for expensive-to-evaluate edges","authors":"Venkatraman Narayanan, Maxim Likhachev","year":"2017","venue":"ICAPS","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Nielsen and Kavraki, 2000] Christian L Nielsen and Ly- dia E Kavraki. A 2 level fuzzy prm for manipulation plan- ning. In IROS, 2000.","title":"Christian L Nielsen and Lydia E Kavraki. A 2 level fuzzy prm for manipulation planning","authors":"Kavraki Nielsen","year":"2000","venue":"IROS","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}