Publications
Dissertation: Scalable Online Decentralized Smoothing and Mapping
Defense: February 2014
Accepted: April 2014
Advisors: Dr. Frank Dellaert and Dr. Ayanna Howard (Co-advisor)
Abstract:
Many applications for field robots can benefit from large numbers of robots, especially applications where the objective is for the robots to cover or explore a region. A key enabling technology for robust autonomy in these teams of small and cheap robots is the development of collaborative perception to account for the shortcomings of the small and cheap sensors on the robots.
In this dissertation, I present DDF-SAM to address the decentralized data fusion (DDF) inference problem with a smoothing and mapping (SAM) approach to single-robot mapping that is online, scalable and consistent while supporting a variety of sensing modalities. The DDF-SAM approach performs fully decentralized simultaneous localization and mapping in which robots choose a relevant subset of variables from their local map to share with neighbors. Each robot summarizes their local map to yield a density on exactly this chosen set of variables, and then distributes this summarized map to neighboring robots, allowing map information to propagate throughout the network. Each robot fuses summarized maps it receives to yield a map solution with an extended sensor horizon.
I introduce two primary variations on DDF-SAM, one that uses a batch nonlinear constrained optimization procedure to combine maps, DDF-SAM 1.0, and one that uses an incremental solving approach for substantially faster performance, DDF-SAM 2.0. I validate these systems using a combination of real-world and simulated experiments. In addition, I evaluate design trade-offs for operations within DDF-SAM, with a focus on efficient approximate map summarization to minimize communication costs.
Journal Papers
Multipolicy decision-making for autonomous driving via changepoint-based behavior prediction: Theory and experiment. Enric Galceran, Alexander G. Cunningham, Ryan M. Eustice and Edwin Olson. Autonomous Robots. August 2017, pp. 1367-1382. [pdf]
Book Chapters
MPDM: Multi-policy Decision-Making from Autonomous Driving to Social Robot Navigation. Alexander G. Cunningham, Enric Galceran, Dhanvin Mehta, Gonzalo Ferrer, Ryan Eustice and Edwin Olson. In Lecture Notes in Control and Information Sciences. Jan 2019, pp. 201-223. [pdf]
Effects of sensory precision on mobile robot localization and mapping. Carlos Nieto-Granda, Alex J. Trevor, John G. Rogers, Alexander G. Cunningham, Manohar Paluri, Nathan Michael, Frank Dellaert, Henrik I.Christensen, and Vijay Kumar. In Experimental Robotics, STAR. Springer Verlag, Heidelberg/New York, Jan 2014, pp. 433-446.
Patents
Systems and methods for vehicular navigation. Alexander G. Cunningham, Robert A.E. Zidek, Noah J. Epstein. Patent #US20200218277A1. Pending. [pdf]
Modeling graph of interactions between agents. Alexander G. Cunningham. Patent #10860025. Granted 2020-12-08. [pdf]
Method and apparatus for road hazard detection. Ryan W. Wolcott, Alexander G. Cunningham. Patent #10754344. Granted 2020-08-25. [pdf]
Vehicle Trajectory Determination. Edwin Olson, Enric Galceran, Alexander G. Cunningham, Ryan M. Eustice, James Robert McBride. Patent #US9934688B2. Granted 2018-04-03. [pdf]
Conference Papers
Continuous-Time Estimation for Dynamic Obstacle Tracking. Arash Ushani, Nicholas Carlevaris-Bianco, Alexander G. Cunningham, Enric Galceran, Ryan M. Eustice. 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (Hamburg, Germany, September 2015).
Cooperative Localization by Factor Composition with Faulty Low-Bandwidth Communication Channels. Jeffrey Walls, Alexander G. Cunningham, Ryan Eustice. 2015 IEEE International Conference on Robotics and Automation (Seattle, USA, May 2015).
[pdf]
Workshop Papers
Understanding the Past to Predict the Future: Multipolicy Decision-Making using Changepoints for Autonomous Driving. Enric Galceran, Alexander G. Cunningham, Ryan M. Eustice and Edwin Olson. ICRA 2015 Workshop: Beyond Geometric Constraints: Planning for Solving Complex Tasks, Reducing Uncertainty, and Generating Informative Paths & Policies (Seattle, Washington, May 2015).
[pdf]
Technical Reports
Local Exponential Maps: Towards Massively Distributed Multi-robot Mapping, Frank Dellaert, Alireza Fathi, Alexander Cunningham, Manohar Paluri, and Kai Ni, GVU Center; College of Computing; Georgia Tech, GIT-GVU-10-04, 2010
EasySLAM, Alireza Fathi, Alexander Cunningham, Manohar Paluri, Kai Ni, and Frank Dellaert, GVU Center; College of Computing; Georgia Tech, GIT-GVU-10-03, 2010