Collaborating Authors


Relative velocity-based reward functions for crowd navigation of robots Artificial Intelligence

The four-wheeled Mecanum robot is widely used in various industries due to its maneuverability and strong load capacity, which is suitable for performing precise transportation tasks in a narrow environment, but while the Mecanum wheel robot has mobility, it also consumes more energy than ordinary robots. The power consumed by the Mecanum wheel mobile robot varies enormously depending on their operating regimes and environments. Therefore, only knowing the working environment of the robot and the accurate power consumption model can we accurately predict the power consumption of the robot. In order to increase the appli-cable scenarios of energy consumption modeling for Mecanum wheel robots and improve the accuracy of energy consumption modeling, this paper focuses on various factors that affect the energy consumption of the Mecanum wheel robot, such as motor temperature, terrain, the center of gravity position, etc. The model is derived from the kinematic and kinetic model combined with electrical engineering and energy flow principles. The model has been simulated in MATLAB and experimentally validated with the four-wheeled Mecanum robot platform in our lab. Experimental results show that the model is 90% accurate. The results of energy consumption modeling can help robots to save energy by helping them to perform rational path planning and task planning.

Exact and Bounded Collision Probability for Motion Planning under Gaussian Uncertainty Artificial Intelligence

Computing collision-free trajectories is of prime importance for safe navigation. We present an approach for computing the collision probability under Gaussian distributed motion and sensing uncertainty with the robot and static obstacle shapes approximated as ellipsoids. The collision condition is formulated as the distance between ellipsoids and unlike previous approaches we provide a method for computing the exact collision probability. Furthermore, we provide a tight upper bound that can be computed much faster during online planning. Comparison to other state-of-the-art methods is also provided. The proposed method is evaluated in simulation under varying configuration and number of obstacles.

Anytime Stochastic Task and Motion Policies Artificial Intelligence

In order to solve complex, long-horizon tasks, intelligent robots need to carry out high-level, abstract planning and reasoning in conjunction with motion planning. However, abstract models are typically lossy and plans or policies computed using them can be inexecutable. These problems are exacerbated in stochastic situations where the robot needs to reason about and plan for multiple contingencies. We present a new approach for integrated task and motion planning in stochastic settings. In contrast to prior work in this direction, we show that our approach can effectively compute integrated task and motion policies whose branching structures encode agent behaviors that handle multiple execution-time contingencies. We prove that our algorithm is probabilistically complete and can compute feasible solution policies in an anytime fashion so that the probability of encountering an unresolved contingency decreases over time. Empirical results on a set of challenging problems show the utility and scope of our method.

Post Triangular Rewiring Method for Shorter RRT Robot Path Planning Artificial Intelligence

This paper proposed the 'Post Triangular Rewiring' method that minimizes the sacrifice of planning time and overcomes the limit of Optimality of sampling-based algorithm such as Rapidly-exploring Random Tree (RRT) algorithm. The proposed 'Post Triangular Rewiring' method creates a closer to the optimal path than RRT algorithm before application through the triangular inequality principle. The experiments were conducted to verify a performance of the proposed method. When the method proposed in this paper are applied to the RRT algorithm, the Optimality efficiency increase compared to the planning time.

Motion Planning for a Pair of Tethered Robots Artificial Intelligence

Considering an environment containing polygonal obstacles, we address the problem of planning motions for a pair of planar robots connected to one another via a cable of limited length. Much like prior problems with a single robot connected via a cable to a fixed base, straight line-of-sight visibility plays an important role. The present paper shows how the reduced visibility graph provides a natural discretization and captures the essential topological considerations very effectively for the two robot case as well. Unlike the single robot case, however, the bounded cable length introduces considerations around coordination (or equivalently, when viewed from the point of view of a centralized planner, relative timing) that complicates the matter. Indeed, the paper has to introduce a rather more involved formalization than prior single-robot work in order to establish the core theoretical result -- a theorem permitting the problem to be cast as one of finding paths rather than trajectories. Once affirmed, the planning problem reduces to a straightforward graph search with an elegant representation of the connecting cable, demanding only a few extra ancillary checks that ensure sufficiency of cable to guarantee feasibility of the solution. We describe our implementation of A${}^\star$ search, and report experimental results. Lastly, we prescribe an optimal execution for the solutions provided by the algorithm.

Neural-Swarm2: Planning and Control of Heterogeneous Multirotor Swarms using Learned Interactions Artificial Intelligence

We present Neural-Swarm2, a learning-based method for motion planning and control that allows heterogeneous multirotors in a swarm to safely fly in close proximity. Such operation for drones is challenging due to complex aerodynamic interaction forces, such as downwash generated by nearby drones and ground effect. Conventional planning and control methods neglect capturing these interaction forces, resulting in sparse swarm configuration during flight. Our approach combines a physics-based nominal dynamics model with learned Deep Neural Networks (DNNs) with strong Lipschitz properties. We evolve two techniques to accurately predict the aerodynamic interactions between heterogeneous multirotors: i) spectral normalization for stability and generalization guarantees of unseen data and ii) heterogeneous deep sets for supporting any number of heterogeneous neighbors in a permutation-invariant manner without reducing expressiveness. The learned residual dynamics benefit both the proposed interaction-aware multi-robot motion planning and the nonlinear tracking control designs because the learned interaction forces reduce the modelling errors. Experimental results demonstrate that Neural-Swarm2 is able to generalize to larger swarms beyond training cases and significantly outperforms a baseline nonlinear tracking controller with up to three times reduction in worst-case tracking errors.

CAMPs: Learning Context-Specific Abstractions for Efficient Planning in Factored MDPs Artificial Intelligence

Meta-planning, or learning to guide planning from experience, is a promising approach to improving the computational cost of planning. A general meta-planning strategy is to learn to impose constraints on the states considered and actions taken by the agent. We observe that (1) imposing a constraint can induce context-specific independences that render some aspects of the domain irrelevant, and (2) an agent can take advantage of this fact by imposing constraints on its own behavior. These observations lead us to propose the context-specific abstract Markov decision process (CAMP), an abstraction of a factored MDP that affords efficient planning. We then describe how to learn constraints to impose so the CAMP optimizes a trade-off between rewards and computational cost. Our experiments consider five planners across four domains, including robotic navigation among movable obstacles (NAMO), robotic task and motion planning for sequential manipulation, and classical planning. We find planning with learned CAMPs to consistently outperform baselines, including Stilman's NAMO-specific algorithm. Video:

Autonomy and Unmanned Vehicles Augmented Reactive Mission-Motion Planning Architecture for Autonomous Vehicles Artificial Intelligence

Advances in hardware technology have facilitated more integration of sophisticated software toward augmenting the development of Unmanned Vehicles (UVs) and mitigating constraints for onboard intelligence. As a result, UVs can operate in complex missions where continuous trans-formation in environmental condition calls for a higher level of situational responsiveness and autonomous decision making. This book is a research monograph that aims to provide a comprehensive survey of UVs autonomy and its related properties in internal and external situation awareness to-ward robust mission planning in severe conditions. An advance level of intelligence is essential to minimize the reliance on the human supervisor, which is a main concept of autonomy. A self-controlled system needs a robust mission management strategy to push the boundaries towards autonomous structures, and the UV should be aware of its internal state and capabilities to assess whether current mission goal is achievable or find an alternative solution. In this book, the AUVs will become the major case study thread but other cases/types of vehicle will also be considered. In-deed the research monograph, the review chapters and the new approaches we have developed would be appropriate for use as a reference in upper years or postgraduate degrees for its coverage of literature and algorithms relating to Robot/Vehicle planning, tasking, routing, and trust.

Deployment and Evaluation of a Flexible Human-Robot Collaboration Model Based on AND/OR Graphs in a Manufacturing Environment Artificial Intelligence

The Industry 4.0 paradigm promises shorter development times, increased ergonomy, higher flexibility, and resource efficiency in manufacturing environments. Collaborative robots are an important tangible technology for implementing such a paradigm. A major bottleneck to effectively deploy collaborative robots to manufacturing industries is developing task planning algorithms that enable them to recognize and naturally adapt to varying and even unpredictable human actions while simultaneously ensuring an overall efficiency in terms of production cycle time. In this context, an architecture encompassing task representation, task planning, sensing, and robot control has been designed, developed and evaluated in a real industrial environment. A pick-and-place palletization task, which requires the collaboration between humans and robots, is investigated. The architecture uses AND/OR graphs for representing and reasoning upon human-robot collaboration models online. Furthermore, objective measures of the overall computational performance and subjective measures of naturalness in human-robot collaboration have been evaluated by performing experiments with production-line operators. The results of this user study demonstrate how human-robot collaboration models like the one we propose can leverage the flexibility and the comfort of operators in the workplace. In this regard, an extensive comparison study among recent models has been carried out.

Learning compositional models of robot skills for task and motion planning Artificial Intelligence

The objective of this work is to augment the basic abilities of a robot by learning to use new sensorimotor primitives to solve complex long-horizon manipulation problems. This requires flexible generative planning that can combine primitive abilities in novel combinations and thus generalize across a wide variety of problems. In order to plan with primitive actions, we must have models of the preconditions and effects of those actions: under what circumstances will executing this primitive successfully achieve some particular effect in the world? We use, and develop novel improvements on, state-of-the-art methods for active learning and sampling. We use Gaussian process methods for learning the conditions of operator effectiveness from small numbers of expensive training examples. We develop adaptive sampling methods for generating a comprehensive and diverse sequence of continuous parameter values (such as pouring waypoints for a cup) configurations and during planning for solving a new task, so that a complete robot plan can be found as efficiently as possible. We demonstrate our approach in an integrated system, combining traditional robotics primitives with our newly learned models using an efficient robot task and motion planner. We evaluate our approach both in simulation and in the real world through measuring the quality of the selected pours and scoops. Finally, we apply our integrated system to a variety of long-horizon simulated and real-world manipulation problems.