Autonomous Robots

This work presents Object Landmarks, a new type of visual feature designed for visual localization over major changes in distance and scale. An Object Landmark consists of a bounding box $${\mathbf {b}}$$ defining an object, a descriptor $${\mathbf {q}}$$ of that object produced by a Convolutional Neural Network, and a set of classical point features within $${\mathbf {b}}$$ . We evaluate Object Landmarks on visual odometry and place-recognition tasks, and compare them against several modern approaches. We find that Object Landmarks enable superior localization over major scale changes, reducing error by as much as 18% and increasing robustness to failure by as much as 80% versus the state-of-the-art. They allow localization under scale change factors up to 6, where state-of-the-art approaches break down at factors of 3 or more.

Trong bài báo này, chúng tôi xem xét việc phát triển một khung điều khiển tương thích động học cho một hệ thống mô-đun gồm các bộ thao tác di động có bánh xe, có khả năng hợp tác để vận chuyển một tải trọng chung. Mỗi bộ manipulator di động độc lập bao gồm một Robot Di động có Bánh (WMR) truyền động khác nhau với một cánh tay manipulator thụ động, hai bậc tự do (d.o.f) có khớp quay gắn trên. Phương tiện di động tổng hợp, được tạo thành bằng cách đặt một tải trọng ở cuối hiệu ứng của hai (hoặc nhiều hơn) bộ manipulators di động như vậy, có khả năng điều chỉnh, phát hiện và sửa lỗi cấu hình tương đối, cả tức thời lẫn hữu hạn. Khung điều khiển/lập kế hoạch chuyển động tương thích động học được phát triển ở đây nhằm duy trì tất cả các ràng buộc động học (holonomic và nonholonomic) trong các hệ thống như vậy. Cho một quỹ đạo hiệu ứng đầu tùy ý, bộ điều khiển phân cấp hai cấp của từng bộ manipulator di động trước tiên tạo ra một quỹ đạo mong muốn phù hợp về động học cho cơ sở WMR, sau đó được theo dõi bởi một bộ điều khiển ổn định tư thế cấp dưới thích hợp. Hai biến thể của các phương án điều khiển hợp tác cấp hệ thống—điều khiển theo mô hình lãnh đạo và điều khiển phân cấp—sau đó được tạo ra dựa trên phương án điều khiển của từng bộ manipulator di động. Cả hai phương pháp đều được đánh giá trong một khung triển khai nhấn mạnh cả mô phỏng ảo (VP) và thí nghiệm phần cứng trong vòng lặp (HIL). Kết quả mô phỏng và thực nghiệm của một ví dụ về hệ thống hai mô-đun được sử dụng để làm nổi bật khả năng của bộ điều khiển dựa trên cảm biến cục bộ theo thời gian thực để điều chỉnh, phát hiện và sửa lỗi hình thành tương đối.

Cleaning is a major problem associated with pools. Since the manual cleaning is tedious and boring there is an interest in automating the task. This paper presents methods for autonomous localization and navigation for a pool cleaner to enable full coverage of pools. Path following cannot be ensured through use of internal position estimation methods alone; therefore sensing is needed. Sensor based estimation enable automatic correction of slippage. For this application we use ultrasonic sonars. Based on an analysis of the overall task and performance of the system a strategy for cleaning/navigation is developed. For the automatic localization a Kalman filtering technique is proposed: the Kalman filter uses sonar measurements and a dynamic model of the robot to provide estimates of the pose of the pool cleaner. Using this localization method we derive an optimal control strategy for traversal of a pool. The system has been implemented and successfully tested on the “WEDAB400” pool cleaner.

For many tasks, we wish to record or recover the description of a remote environment so that it can be inspected by a person. This is the problem we address in this paper. Rather than recovering a geometric description of an environment, as many robotics systems attempt to do, we seek to recover a model of an environment in terms of its appearance from a set of carefully selected viewpoints. Our hope is that this type of model is both more accessible to humans for many realistic tasks, and also more readily achieved with automated systems. These viewpoints are locations in the environment associated with views containing maximal visual interest. This approach to environment representation is analogous to image compression. Our goal is to obtain a set of representative views resembling those that would be selected by a human observer given the same task. Our computational procedure is inspired by models of human visual attention appearing in the literature on human psychophysics. We make use of the underlying edge structure of a scene, as it is largely unaffected by variations in illumination. Our implementation uses a mobile robot to traverse the environment, and then builds an image-based virtual representation of the environment, only keeping the views whose responses were highest. We demonstrate the effectiveness of our attention operator on both single images, and in viewpoint selection within an unknown environment.

This paper proposes a novel sampling-based motion planner, which integrates in Rapidly exploring Random Tree star ( $$\hbox {RRT}^{\star }$$ ) a database of pre-computed motion primitives to alleviate its computational load and allow for motion planning in a dynamic or partially known environment. The database is built by considering a set of initial and final state pairs in some grid space, and determining for each pair an optimal trajectory that is compatible with the system dynamics and constraints, while minimizing a cost. Nodes are progressively added to the tree of feasible trajectories in the $$\hbox {RRT}^{\star }$$ algorithm by extracting at random a sample in the gridded state space and selecting the best obstacle-free motion primitive in the database that joins it to an existing node. The tree is rewired if some nodes can be reached from the new sampled state through an obstacle-free motion primitive with lower cost. The computationally more intensive part of motion planning is thus moved to the preliminary offline phase of the database construction at the price of some performance degradation due to gridding. Grid resolution can be tuned so as to compromise between (sub)optimality and size of the database. The planner is shown to be asymptotically optimal as the grid resolution goes to zero and the number of sampled states grows to infinity.

Robots acting in human environments usually need to perform multiple motion and force tasks while respecting a set of constraints. When a physical contact with the environment is established, the newly activated force task or contact constraint may interfere with other tasks. The objective of this paper is to provide a control framework that can achieve real-time control of humanoid robots performing both strict and non strict prioritized motion and force tasks. It is a torque-based quasi-static control framework, which handles a dynamically changing task hierarchy with simultaneous priority transitions as well as activation or deactivation of tasks. A quadratic programming problem is solved to maintain desired task hierarchies, subject to constraints. A generalized projector is used to quantitatively regulate how much a task can influence or be influenced by other tasks through the modulation of a priority matrix. By the smooth variations of the priority matrix, sudden hierarchy rearrangements can be avoided to reduce the risk of instability. The effectiveness of this approach is demonstrated on both a simulated and a real humanoid robot.

This paper presents a design of a three-fingered robotic hand driven by active and passive tendons and proposes control methods for this hand. The tendon-driven robotic hand consists of the thumb, the index and the middle fingers. The robotic thumb can move all the joints independently. In contrast, the index and the middle robotic fingers are under-actuated using the combination of active and passive tendons, and move the terminal two joints synchronously, which is one of the important features of the human digits. We present passivity-based impedance and force controllers for tendon-driven robotic fingers and discuss how to combine them for fast and secure grasps. We experimentally validate that the robotic hand moves fast and manipulates an object and demonstrate that the robotic hand grasps objects in diverse ways.