Learning from demonstration (LfD) provides a fast, intuitive and efficient framework to program robot skills, which has gained growing interest both in research and industrial applications. Most complex manipulation tasks are long-term and involve a set of skill primitives. Thus it is crucial to have a reliable coordination scheme that selects the correct sequence of skill primitive and the correct parameters for each skill, under various scenarios. Instead of relying on a precise simulator, this work proposes a human-in-the-loop coordination framework for LfD skills that: builds parameterized skill models from kinesthetic demonstrations; constructs a geometric task network (GTN) on-the-fly from human instructions; learns a hierarchical control policy incrementally during execution. This framework can reduce significantly the manual design efforts, while improving the adaptability to new scenes. We show on a 7-DoF robotic manipulator that the proposed approach can teach complex industrial tasks such as bin sorting and assembly in less than 30 minutes.

Complex manipulation tasks can contain various execution branches of primitive skills in sequence or in parallel under different scenarios. Manual specifications of such branching conditions and associated skill parameters are not only error-prone due to corner cases, but also quickly untraceable given a large number of objects and skills. On the other hand, learning from demonstration has increasingly shown to be an intuitive and effective way to program such skills for industrial robots. Parameterized skill representations allow generalization over new scenarios, which however makes the planning process much slower thus unsuitable for online applications. In this article, we propose a hierarchical and compositional planning framework that learns a geometric task network (GTN) from exhaustive planners, without any manual inputs. A GTN is a goal-dependent task graph that encapsulates both the transition relations among skill representations and the geometric constraints underlying these transitions. This framework has shown to improve dramatically the offline learning efficiency, the online performance, and the transparency of decision process, by leveraging the task-parameterized models. We demonstrate the approach on a 7-DoF robot arm both in simulation and on hardware solving various manipulation tasks.

Dense Object Nets (DONs) by Florence, Manuelli and Tedrake (2018) introduced dense object descriptors as a novel visual object representation for the robotics community. It is suitable for many applications including object grasping, policy learning, etc. DONs map an RGB image depicting an object into a descriptor space image, which implicitly encodes key features of an object invariant to the relative camera pose. Impressively, the self-supervised training of DONs can be applied to arbitrary objects and can be evaluated and deployed within hours. However, the training approach relies on accurate depth images and faces challenges with small, reflective objects, typical for industrial settings, when using consumer grade depth cameras. In this paper we show that given a 3D model of an object, we can generate its descriptor space image, which allows for supervised training of DONs. We rely on Laplacian Eigenmaps (LE) to embed the 3D model of an object into an optimally generated space. While our approach uses more domain knowledge, it can be efficiently applied even for smaller and reflective objects, as it does not rely on depth information. We compare the training methods on generating 6D grasps for industrial objects and show that our novel supervised training approach improves the pick-and-place performance in industry-relevant tasks.

Learning from Demonstration (LfD) provides an intuitive and fast approach to program robotic manipulators. Task parameterized representations allow easy adaptation to new scenes and online observations. However, this approach has been limited to pose-only demonstrations and thus only skills with spatial and temporal features. In this work, we extend the LfD framework to address forceful manipulation skills, which are of great importance for industrial processes such as assembly. For such skills, multi-modal demonstrations including robot end-effector poses, force and torque readings, and operation scene are essential. Our objective is to reproduce such skills reliably according to the demonstrated pose and force profiles within different scenes. The proposed method combines our previous work on task-parameterized optimization and attractor-based impedance control. The learned skill model consists of (i) the attractor model that unifies the pose and force features, and (ii) the stiffness model that optimizes the stiffness for different stages of the skill. Furthermore, an online execution algorithm is proposed to adapt the skill execution to real-time observations of robot poses, measured forces, and changed scenes. We validate this method rigorously on a 7-DoF robot arm over several steps of an E-bike motor assembly process, which require different types of forceful interaction such as insertion, sliding and twisting.

Enabling robots to quickly learn manipulation skills is an important, yet challenging problem. Such manipulation skills should be flexible, e.g., be able adapt to the current workspace configuration. Furthermore, to accomplish complex manipulation tasks, robots should be able to sequence several skills and adapt them to changing situations. In this work, we propose a rapid robot skill-sequencing algorithm, where the skills are encoded by object-centric hidden semi-Markov models. The learned skill models can encode multimodal (temporal and spatial) trajectory distributions. This approach significantly reduces manual modeling efforts, while ensuring a high degree of flexibility and re-usability of learned skills. Given a task goal and a set of generic skills, our framework computes smooth transitions between skill instances. To compute the corresponding optimal end-effector trajectory in task space we rely on Riemannian optimal controller. We demonstrate this approach on a 7 DoF robot arm for industrial assembly tasks.

Guaranteeing safe constraint satisfaction in nonlinear control systems with uncertainty remains a major challenge for control. The most successful control method handling constraints under uncertainty has without doubt been model predictive control (MPC). In particular, recent sampling-based MPC methods have shown success in controlling stochastic systems with complex, nonlinear dynamics. The sampling-based schemes are appealing since they do not need strong assumptions on the underlying model, except that it can be forward simulated. At the same time, the lack of major assumptions on the models make the statement of safety or robustness guarantees difficult. However, the samples drawn during the control process inherently contain probabilistic information about these properties. In this paper, we formally describe the problem that results by adding chance constraints to a sampling-based MPC scheme. Furthermore, based on a variant of the Chernoff bound, we derive trust regions, in which the sampling based estimation of the safety constraint satisfies a specified quality. Finally, we present a case study in the navigation domain to demonstrate the applicability of the proposed approach.

Coordinating a team of robots to fulfill a common task is still a demanding problem. This is even more the case when considering uncertainty in the environment, as well as temporal dependencies within the task specification. A multi-robot cooperation from a single goal specification requires mechanisms for decomposing the goal as well as an efficient planning for the team. However, planning action sequences offline is insufficient in real world applications. Rather, due to uncertainties, the robots also need to closely coordinate during execution and adjust their policies when additional observations are made. The framework presented in this paper enables the robot team to cooperatively fulfill tasks given as temporal logic specifications while explicitly considering uncertainty and incorporating observations during execution. We present the effectiveness of our ROS implementation of this approach in a case study scenario.

We present an efficient approach to plan action sequences for a team of robots from a single finite LTL mission specification. The resulting execution strategy is proven to solve the given mission with minimal team costs, e.g., with shortest execution time. For planning, an established graph-based search method based on the multi-objective shortest path problem is adapted to multi-robot planning and extended to support resource constraints. We further improve planning efficiency significantly for missions which consist of independent parts by using previous results regarding LTL decomposition. The efficiency and practicality of the ROS implementation of our approach is demonstrated in example scenarios.