This paper describes a collection of optimization algorithms for achieving dynamic planning, control, and state estimation for a bipedal robot designed to operate reliably in complex environments. To make challenging locomotion tasks tractable, we describe several novel applications of convex, mixed-integer, and sparse nonlinear optimization to problems ranging from footstep placement to whole-body planning and control. We also present a state estimator formulation that, when combined with our walking controller, permits highly precise execution of extended walking plans over non-flat terrain. We describe our complete system integration and experiments carried out on Atlas, a full-size hydraulic humanoid robot built by Boston Dynamics, Inc.

We present a new SLAM system capable of producing high quality globally consistent surface reconstructions over hundreds of metres in real-time with only a low-cost commodity RGB-D sensor. By using a fused volumetric surface reconstruction we achieve a much higher quality map over what would be achieved using raw RGB-D point clouds. In this paper we highlight three key techniques associated with applying a volumetric fusion-based mapping system to the SLAM problem in real-time. First, the use of a GPU-based 3D cyclical buffer trick to efficiently extend dense every frame volumetric fusion of depth maps to function over an unbounded spatial region. Second, overcoming camera pose estimation limitations in a wide variety of environments by combining both dense geometric and photometric camera pose constraints. Third, efficiently updating the dense map according to place recognition and subsequent loop closure constraints by the use of an "as-rigid-as-possible" space deformation. We present results on a wide variety of aspects of the system and show through evaluation on de facto standard RGB-D benchmarks that our system performs strongly in terms of trajectory estimation, map quality and computational performance in comparison to other state-of-the-art systems.

In this paper we propose a method for robust dense RGB-D SLAM in dynamic environments which detects moving objects and simultaneously reconstructs the background structure. Dynamic environments are challenging for visual SLAM as moving objects can impair camera pose tracking and cause corruptions to be integrated into the map. While most methods employ implicit robust penalizers or outlier filtering techniques in order to handle moving objects, our approach is to simultaneously estimate the camera motion as well as a probabilistic static/dynamic segmentation of the current RGB-D image pair. This segmentation is then used for weighted dense RGB-D fusion to estimate a 3D model of only the static parts of the environment. By leveraging the 3D model for frameto-model alignment, as well as static/dynamic segmentation, camera motion estimation has reduced overall drift -as well as being more robust to the presence of dynamics in the scene. Demonstrations are presented which compare the proposed method to comparable state-of-the-art approaches using both static and dynamic sequences. The proposed method achieves similar performance in static environments and improved accuracy and robustness in dynamic scenes.

Self-localization of an underwater vehicle is particularly challenging due to the absence of Global Positioning System (GPS) reception or features at known positions that could otherwise have been used for position computation. Thus Autonomous Underwater Vehicle (AUV) applications typically require the pre-deployment of a set of beacons.This thesis examines the scenario in which the members of a group of AUVs exchange navigation information with one another so as to improve their individual position estimates.We describe how the underwater environment poses unique challenges to vehicle navigation not encountered in other environments in which robots operate and how cooperation can improve the performance of self-localization. As intra-vehicle communication is crucial to cooperation, we also address the constraints of the communication channel and the effect that these constraints have on the design of cooperation strategies.The classical approaches to underwater self-localization of a single vehicle, as well as more recently developed techniques are presented. We then examine how methods used for cooperating land-vehicles can be transferred to the underwater domain. An algorithm for distributed self-localization, which is designed to take the specific characteristics of the environment into account, is proposed.We also address how correlated position estimates of cooperating vehicles can lead to overconfidence in individual position estimates.Finally, key to any successful cooperative navigation strategy is the incorporation of the relative positioning between vehicles. The performance of localization algorithms with different geometries is analyzed and a distributed algorithm for the dynamic positioning of vehicles, which serve as dedicated navigation beacons for a fleet of AUVs, is proposed.

This paper describes the experimental implementation of an online algorithm for cooperative localization of submerged autonomous underwater vehicles (AUVs) supported by an autonomous surface craft. Maintaining accurate localization of an AUV is difficult because electronic signals, such as GPS, are highly attenuated by water. The usual solution to the problem is to utilize expensive navigation sensors to slow the rate of dead-reckoning divergence. We investigate an alternative approach that utilizes the position information of a surface vehicle to bound the error and uncertainty of the on-board position estimates of a low-cost AUV. This approach uses the Woods Hole Oceanographic Institution (WHOI) acoustic modem to exchange vehicle location estimates while simultaneously estimating inter-vehicle range. A study of the system observability is presented so as to motivate both the choice of filtering approach and surface vehicle path planning. The first contribution of this paper is the presentation of an experiment in which an extended Kalman filter (EKF) implementation of the concept ran online on-board an OceanServer Iver2 AUV while supported by an autonomous surface vehicle moving adaptively. The second contribution of this paper is provide a quantitative performance comparison of three estimators: particle filtering (PF), Nonlinear Least Squares optimization (NLS), and the EKF for a mission using three autonomous surface craft (two operating in the AUV role). Our results indicate that the PF and NLS estimators outperform the EKF, with NLS providing the best performance.

Abstract-In this paper, we demonstrate a system for temporally scalable visual SLAM using a reduced pose graph representation. Unlike previous visual SLAM approaches that use keyframes, our approach continually uses new measurements to improve the map, yet achieves efficiency by avoiding adding redundant frames and not using marginalization to reduce the graph. To evaluate our approach, we present results using an online binocular visual SLAM system that uses place recognition for both robustness and multi-session operation. To allow largescale indoor mapping, our system automatically handles elevator rides based on accelerometer data. We demonstrate long-term mapping in a large multi-floor building, using approximately nine hours of data collected over the course of six months. Our results illustrate the capability of our visual SLAM system to scale in size with the area of exploration instead of the time of exploration.

The DARPA Robotics Challenge Trials held in December 2013 provided a landmark demonstration of dexterous mobile robots executing a variety of tasks aided by a remote human operator using only data from the robot's sensor suite transmitted over a constrained, fieldrealistic communications link. We describe the design considerations, architecture, implementation and performance of the software that Team MIT developed to command and control an Atlas humanoid robot. Our design emphasized human interaction with an efficient motion planner, where operators expressed desired robot actions in terms of affordances fit using perception and manipulated in a custom user interface. We highlight several important lessons we learned while developing our system on a highly compressed schedule.

No abstract