This paper explores the challenge of optimal routing for a mobile robot navigating a dynamic and shared human environment. The primary goal is to minimize the risk of performance degradation during motion, such as delays in completing tasks due to the need for safe or acceptable human robot encounters. The problem is formulated as a graph whose edge costs become progressively known only as the robot moves through the environment. We model this problem as a Markov Decision Process (MDP), enabling an offline evaluation of the expected cost of alternative routes based on statistical information about human spatial distributions and possible observations at each intersection. This compact state representation scales linearly with the number of intersections in the map. Since the memoryless property of the MDP may induce loops during online execution, we compute an offline policy and introduce an online policy adaptation mechanism to prevent cyclic behaviors. Exten sive simulations across environments of different complexity, and using data collected from real-world experiments, demonstrate that our approach outperforms reactive and advanced state-of the-art planners in terms of either performance or scalability.

Due to their fast and efficient locomotion, two-wheeled humanoids are fascinating systems with the potential to be involved in many application domains, including healthcare, manufacturing, and many others. However, these robots constitute a challenging case of study for control purposes due to the two-wheeled inverted pendulum dynamics that characterizes their mobility and support, as it is underactuated and unstable. In this article, we propose a novel whole-body control approach to stabilize two-wheeled humanoids. To tackle the control problem of their forward motion and pitch equilibrium, leveraging on the observation that such systems are usually characterized by a faster and a slower dynamics (being the pitch angle faster and the forward displacement slower), we design a composite whole-body control that combines two computed-torque control loops to stabilize both dynamics to the desired trajectories. The control approach is introduced and its derivation is described for the simpler case of a two-wheeled inverted pendulum first, and for a whole two-wheeled humanoid after. To prove its validity, the control approach is tested experimentally on the two-wheeled humanoid robot Alter-Ego. The robot proves to be able to perform complicated interaction tasks, including opening a door, grasping a heavy object, and resisting to external dynamic disturbances.

In recent years, robotics has been largely applied to improve the efficiency of logistic processes. Pallets cover a crucial role in the logistic flow, since they represent the main way to store and ship items. When put onto pallets, the items are wrapped with plastic films to protect them and prevent them from falling. Despite being the first and necessary operation for handling the stacked goods, unwrapping—the task of removing the plastic films wrapped around the goods—has not yet been satisfactorily automated. We propose the first robotic solution for autonomous unwrapping of generally shaped pallets, including both homogeneous and heterogeneous pallets. Force and torque measurements are exploited to retrieve information on the collisions between the end-effector and the wrapped items or the plastic film. Based on the contact information, we design a novel reactive planning strategy that makes the unwrapping task effective and robust on pallets with uncertain position or shape. We present the results of an extensive experimental campaign to validate the proposed method. Note to Practitioners—This work is motivated by the fact that unwrapping machines are not yet common on the market. The few commercial examples are usually bulky machines that lack the flexibility to adapt to different and irregularly shaped pallets. Thus, the crucial operation of removing the plastic film around palletized goods is still mainly performed by hand. Blade handling, ladders, and electrostatic shocks are sources of potential injury. We propose a flexible, autonomous unwrapping robot suitable for both cuboid and irregularly shaped pallets. The robot is composed of a robotic arm, a custom cutting end-effector, a vision module, and a suitable planning and control unit. The reduced dimensions allow it to be mounted on a mobile base. The robot has been successfully tested on different pallet configurations. However, extensive testing in real-world scenarios should be carried out to assess both reliability and time efficiency in more realistic working conditions. Moreover, real pallets can reach considerable heights. Thus, a prismatic joint should be integrated to address such cases. Finally, unwrapping in the presence of typical plastic straps and different types of film, e.g., the shrink one, is to be evaluated.

With the improvements in their computational and physical intelligence, robots are now capable of operating in real-world environments. However, manipulation and grasping capabilities are still areas that require significant improvements. To address this, we introduce a new data-driven grasp planning algorithm called Grasp it Like a Pro 2.0. This algorithm utilizes a small number of human demonstrations to teach a robot how to grasp arbitrary objects. By decomposing objects into basic shapes, our algorithm generates candidate grasps that can generalize to different object's geometry. The algorithm selects the grasp to execute based on a selection policy that maximizes a novel grasp quality metric introduced in this article. This metric considers the complex interdependencies between the predicted grasp, the local approximation produced by the basic shape decomposition, and the gripper used. We evaluate our approach against multiple baselines using different grippers and objects. The results demonstrate the effectiveness of our method in generating and selecting high-quality and reliable grasps. With a soft underactuated robotic hand, our algorithm achieves a 94.0% success rate in 150 grasps across 30 different objects. Similarly, with a rigid gripper, it achieves an 85.0% success rate in 80 grasps across 16 different objects.

Long-time operations of autonomous vehicles and mobile robots in logistics and service applications are still a challenge. To avoid a continuous re-mapping, the map can be updated to obtain a consistent representation of the current environment. In this paper, we propose a novel LIDAR-based occupancy grid map updating algorithm for dynamic environments taking into account possible localisation and measurement errors. The proposed approach allows robust long-term operations as it can detect changes in the working area even in presence of moving elements. Results highlighting map quality and localisation performance, both in simulation and experiments, are reported.

This paper proposes an online optimal perception-aware strategy meant to maximize the information collected along the trajectory via the available measurements while simultaneously minimizing the negative effects of actuation/process noise. Indeed, in several robotic applications, the actuation/process noise is far from negligible and its negative effects are particularly relevant especially with intermittent measurements (e.g. collected by a vision system with limited Field-Of-View). New metrics are proposed as combinations of the Constructability Gramian, for measuring the amount of information collected via the available sensors, and the Reachability Gramian, for measuring the degrading effects of actuation/process noise. Control inputs that optimize those cost functions are provided. To show the effectiveness of our method, we consider one case study involving a unicycle-like vehicle, subject to Gaussian measurement noise and Gaussian or Brownian actuation noise, that estimates its state using intermittent distances from known environmental markers.

Formation control of multi-robot systems has been largely studied due to its wide application domain. Several methods in the literature rely on explicit communication among the robots, which in realistic scenarios may lead to reduced performance or even instability due to delays and packet loss or corruption. Nonetheless, multi-robot coordination based solely on implicit communication has been proposed in cooperative manipulation problems. Taking inspiration from this, we propose a method to solve the formation control problem for a group of ground robots not relying on direct communication among them. Instead, the robots are physically constrained to a common object through elastic cables in order to exploit forces as a means of indirect communication. After deriving the dynamic equations, the control and planning approaches are explained, and the stability of the controlled system is discussed using Lyapunov’s stability theory. Numerical simulations are presented to support the method.

Multirobot fleets play an important role in industrial logistics, surveillance, and exploration applications. A wide literature exists on the topic, both resorting to reactive (i.e. collision avoidance) and to deliberative (i.e. motion planning) techniques. In this work, null space-based inverse kinematics (NSB-IK) methods are applied to the problem of fleet management. Several NSB-IK approaches existing in the literature are reviewed, and compared with a reverse priority approach, which originated in manipulator control, and is here applied for the first time to the considered problem. All NSB-IK approaches are here described in a unified formalism, which allows (i) to encode the property of each controller into a set of seven main key features, (ii) to study possible new control laws with an opportune choice of these parameters. Furthermore, motivated by the envisioned application scenario, we tackle the problem of task-switching activation. Leveraging on the iCAT TPC technique Simetti and Casalino, 2016, in this article, we propose a method to obtain continuity in the control in face of activation or deactivation of tasks, and subtasks by defining suitable damped projection operators. The proposed approaches are evaluated formally, and via simulations. Performances with respect to standard methods are compared considering a specific case study for multivehicles management.

A standing challenge in multirobot systems is to realize safe and efficient motion planning and coordination methods that are capable of accounting for uncertainties and contingencies. The challenge is rendered harder by the fact that robots may be heterogeneous and that their plans may be posted asynchronously. Most existing approaches require constraints on the infrastructure or unrealistic assumptions on robot models. In this article, we propose a centralized, loosely-coupled supervisory controller that overcomes these limitations. The approach responds to newly posed constraints and uncertainties during trajectory execution, ensuring at all times that planned robot trajectories remain kinodynamically feasible, that the fleet is in a safe state, and that there are no deadlocks or livelocks. This is achieved without the need for hand-coded rules, fixed robot priorities, or environment modification. We formally state all relevant properties of robot behavior in the most general terms possible, without assuming particular robot models or environments, and provide both formal and empirical proof that the proposed fleet control algorithms guarantee safety and liveness.

Aerial robotic manipulation is an emergent trend that poses several challenges. To overcome some of these, the DLR cable-Suspended Aerial Manipulator (SAM) has been envisioned. SAM is composed of a fully actuated multi-rotor anchored to a main carrier through a cable and a KUKA LWR attached below the multi-rotor. This work presents a control method to allow SAM, which is a holonomically constrained system, to perform such interaction tasks using a hierarchical control framework. Within this framework, compliance control of the manipulator end-effector is considered to have the highest priority. The second priority is the control of the oscillations induced by, for example, the motion of the arm or physical contact with the environment. A third priority task is related to the internal motion of the manipulator. The proposed approach is validated through simulations and experiments.

Autonomous vehicles are undergoing a rapid development thanks to advances in perception, planning and control methods and technologies achieved in the last two decades. Moreover, the lowering costs of sensors and computing platforms are attracting industrial entities, empowering the integration and development of innovative solutions for civilian use. Still, the development of autonomous racing cars has been confined mainly to laboratory studies and small to middle scale vehicles. This paper tackles the development of a planning and control framework for an electric full scale autonomous racing car, which is an absolute novelty in the literature, upon which we report our preliminary experiments and perspectives on future work. Our system leverages real time Nonlinear Model Predictive Control to track a pre-planned racing line. We describe the whole control system architecture including the mapping and localization methods employed.

In this paper we present a novel geometric approach to motion planning for constrained robot systems. This problem is notoriously hard, as classical sampling-based methods do not easily apply when motion is constrained in a zero-measure submanifold of the configuration space. Based on results on the functional controllability theory of dynamical systems, we obtain a description of the complementary spaces where rigid body motions can occur, and where interaction forces can be generated, respectively. Once this geometric setting is established, the motion planning problem can be greatly simplified. Indeed, we can relax the geometric constraint, i.e., replace the lower-dimensional constraint manifold with a full-dimensional boundary layer. This in turn allows us to plan motion using state-of-the-art methods, such as RRT*, on points within the boundary layer, which can be efficiently sampled. On the other hand, the same geometric approach enables the design of a completely decoupled control scheme for interaction forces, so that they can be regulated to zero (or any other desired value) without interacting with the motion plan execution. A distinguishing feature of our method is that it does not use projection of sampled points on the constraint manifold, thus largely saving in computational time, and guaranteeing accurate execution of the motion plan. An explanatory example is presented, along with an experimental implementation of the method on a bimanual manipulation workstation.

The purpose of this work is to move a step toward the automation of industrial plants through full exploitation of autonomous robots. A planning algorithm is proposed to move different objects in desired configurations with heterogeneous robots such as manipulators, mobile robots and conveyor belts. The proposed approach allows different objects to be handled by different robots simultaneously in an efficient way and avoiding collisions with the environment and self-collisions between robots. In particular, the integrated system will be capable of planning paths for a set of objects from various starting points in the environment (e.g. shelves) to their respective final destinations. The proposed approach unifies the active (e.g., grasping by a hand) and passive (e.g., holding by a table) steps involved in moving the objects in the environment by treating them as end-effectors with constraints and capabilities. Time varying graphs will be introduced to model the problem for simultaneous handling of objects by different end-effectors. Optimal exploration of such graphs will be used to determine paths for each object with time constraints. Results will be validated through simulations.

Random sampling-based methods for motion planning of constrained robot manipulators have been widely studied in recent years. The main problem to deal with is the lack of an explicit parametrization of the non linear submanifold in the Configuration Space (CS) imposed by the constraints in the system. Most of the proposed planning methods use projections to generate valid configurations of the system slowing the planning process. Recently, new robot mechanism includes compliance either in the structure or in the controllers. In this kind of robot most of the times the planned trajectories are not executed exactly due to uncertainties and interactions with the environment. Indeed, controller references are generated such that the constraint is violated to indirectly generate forces during interactions. With the purpose of avoiding projections, in this paper we take advantage of the compliance of systems to relax the geometric constraints imposed by closed kinematic chains. The relaxed constraint is then used in a state-of-the-art suboptimal random sampling based technique to generate paths for constrained robot manipulators. As a consequence of relaxation, arising contact forces acting on the constraint change from configuration to configuration during the planned path. Those forces can be regulated using a proper controller that takes advantage of the geometric decoupling of the subspaces describing constrained rigid-body motions of the mechanism and the controllable forces.

In this paper a model-free approach for damping control of Variable Stiffness Actuators is proposed. The idea is to take advantage of the possibility to change the stiffness of the actuators in controlling the damping. The problem of minimizing the terminal energy for a one degree of freedom spring-mass model with controlled stiffness is first considered. The optimal bang-bang control law uses a maximum stiffness when the link gets away from the desired position, i.e. the link velocity is decreasing, and a minimum one when the link is going towards it, i.e. the link velocity is increasing. Based on Lyapunov stability theorems the obtained law has been proved to be stable for a multi-DoF system. Finally, the proposed control law has been tested and validated through experimental tests.

In this paper, we study the minimum length paths covered by the center of a unicycle equipped with a limited field-of-view (FOV) camera, which must keep a given landmark in sight. Previous works on this subject have provided the optimal synthesis for the cases in which the FOV is only limited in the horizontal directions (i.e., left and right bounds) or in the vertical directions (i.e., upper and lower bounds). In this paper, we show how to merge previous results and hence obtain, for a realistic image plane modeled as a rectangle, a finite alphabet of extremal arcs and the overall synthesis. As shown, this objective cannot be straightforwardly achieved from previous results but needs further analysis and developments. Moreover, there are initial configurations such that there exists no optimal path. Nonetheless, we are always able to provide an $\varepsilon$-optimal path whose length approximates arbitrarily well any other shorter path. As final results, we provide a partition of the motion plane in regions such that the optimal or $\varepsilon$ -optimal path from each point in that region is univocally determined.

In this paper we propose a distributed coordination algorithm for safe and efficient traffic management of heterogeneous robotic agents, moving within dynamic large scale industrial environments. The algorithm consists of a distributed resource-sharing protocol involving a re-planning strategy. Once every agent is assigned with a desired motion path, the algorithm ensures ordered traffic flows of agents, that avoid inter-robot collision and system deadlock (stalls). The algorithm allows multi-level representation of the environment, i.e. large or complex rooms may be seen as a unique resource with given capacity at convenience, which makes the approach appealing for complex industrial environments. Under a suitable condition on the maximum number of agents with respect to the capacity of the environment, we prove that the algorithm correctly allows mutual access to shared resources while avoiding deadlocks. The proposed solution requires no centralized mechanism, no shared memory or ground infrastructure support. Only a local inter-robot communication is required, i.e. every agent must communicate with a limited number of other spatially adjacent robots. We finally show the effectiveness of the proposed approach by simulations, with application to an industrial scenario.

In this paper we consider the motion planning problem for a 3D-Dubins system consisting of a pair of Dubins vehicles moving in a 3D space while maintaining constant distance. We provide a motion planning algorithm and sufficient conditions on the initial and final configurations that guarantee the existence of admissible controls, moving a first step towards the complete controllability characterization of such type of systems. Results obtained in this paper are particularly relevant in order to solve formation control problem for multiple robots as aerial or underwater vehicles, which move in 3D spaces. Simulation results demonstrate the feasibility of the motion planning algorithm proposed in this paper.

We present a complete characterization of shortest paths to a goal position for a vehicle with unicycle kinematics and a limited range sensor, constantly keeping a given landmark in sight. Previous work on this subject studied the optimal paths in case of a frontal, symmetrically limited Field-Of-View (FOV). In this paper we provide a generalization to the case of arbitrary FOVs, including the case that the direction of motion is not an axis of symmetry for the FOV, and even that it is not contained in the FOV. The provided solution is of particular relevance to applications using side-scanning, such as e.g. in underwater sonar-based surveying and navigation.

This paper presents an optimal feedback control scheme to drive a vehicle equipped with a limited Field-Of-View (FOV) camera towards a desired position following the shortest path and keeping a given landmark in sight. Based on the shortest path synthesis available from previous works, feedback control laws are defined for any point on the motion plane exploiting geometric properties of the synthesis itself. Moreover, by using a slightly generalized stability analysis setting, which is that of stability on a manifold, a proof of stability is given. Reported simulations demonstrate the effectiveness of the proposed technique.

Decentralized coordination of multi-agents requires that every agent reliably and efficiently disseminates its state to neighbours through a wireless network. If dissemination is unreliable, safety issues may ensue. Unfortunately, the broadcast service of wireless network is efficient but unreliable (e.g., IEEE 802.11). The Neighbourhood Monitoring Protocol (NMP) is an efficient and scalable protocol that assures a reliable state dissemination between mobile agents, under some conditions of channel utilization. NMP runs on top of IEEE 802.11. In this paper we evaluate NMP with a specific decentralized collision avoidance algorithm based on the GRP policy. The algorithm is particularly challenging because it accommodates an arbitrary number non-holonomic agents. We show that NMP allows the system to scale well and provides a very high state delivery ratio even if it operates on the unreliable broadcast service like 802.11. Doing so, NMP assures the correct state information to the collision avoidance algorithm.

This paper studies the controllability of pairs of identical nonholonomic vehicles maintaining a constant distance. The study provides controllability results for the five most common types of robot vehicles: Dubins, Reeds-Shepp, differential drive, car-like and convexified Reeds-Shepp. The challenge of achieving controllability of such systems is that their admissible control domains depend on configuration variables. A theorem of controllability specifical for such systems has been obtained based on known controllability theorems. As a result, we show that pairs of the latter three types are completely controllable, i.e. can be steered between any two arbitrary configurations. The same does not hold for pairs of Dubins or Reeds-Shepp vehicles, and a description of the reachable sets in these cases is provided. Finally, as direct extension of controllability results of pairs of identical vehicles, the controllability results for two kinds of formation of n identical vehicles are presented.

This paper presents a complete characterization of shortest paths to a goal position for a robot with unicycle kinematics and an on-board camera with limited field-of-view (FOV), which must keep a given feature in sight. Previous work on this subject has shown that the search for a shortest path can be limited to simple families of trajectories. In this paper, we provide a complete optimal synthesis for the problem, i.e., a language of optimal control words, and a global partition of the motion plane induced by shortest paths, such that a word in the optimal language is univocally associated with a region and completely describes the shortest path from any starting point in that region to the goal point. An efficient algorithm to determine the region in which the robot is at any time is also provided.

In this paper we deal with the optimal feedback synthesis problem for robotic vehicles with trailers which can be modeled by differential equations in chained-form. With respect to classical methods for numerical evolution of optimal feedback synthesis via dynamic programming which are based on both input and state discretization, our method exploits the lattice structure naturally imposed on the reachable set by input quantization. A generalized Dijkstra algorithm can be used to obtain sub-optimal (optimal up to the lattice resolution) feedback laws, for chained-form vehicles with n-trailers, in an effective way.

In this paper, we propose a novel policy for steering multiple vehicles between assigned start and goal configurations, ensuring collision avoidance. The policy rests on the assumption that all agents are cooperating by implementing the same traffic rules. However, the policy is completely decentralized, as each agent decides its own motion by applying those rules only on the locally available information, and scalable, in the sense that the amount of information processed by each agent and the computational complexity of the algorithms do not increase with the number of agents in the scenario. The proposed policy applies to systems in which new vehicles may enter the scene and start interacting with existing ones at any time, while others may leave. Under mild conditions on the initial configurations, the policy is shown to be safe, i.e., it guarantees collision avoidance throughout the system evolution. In the paper, conditions are discussed on the desired configurations of agents, under which the ultimate convergence of all vehicles to their goals can also be guaranteed. To show that such conditions are actually necessary and sufficient, which turns out to be a challenging liveness-verification problem for a complex hybrid automaton, we employ a probabilistic verification method. The paper finally presents and discusses simulations for systems of several tens of vehicles, and reports on some experimental implementation showing the practicality of the approach.

In this paper, we consider a decentralized cooperative control policy proposed recently for steering multiple nonholonomic vehicles between assigned start and goal configurations while avoiding collisions. The policy is known to ensure safety (i.e., collision avoidance) for an arbitrarily large number of vehicles, if initial configurations satisfy certain conditions. The method is highly scalable, and effective solutions can be obtained for several tens of autonomous agents. On the other hand, the liveness properties of the policy, i.e. the capability of negotiating a solution in finite time, are not completely understood yet. In this paper, we introduce a condition on the final vehicle configurations, which we conjecture to be necessary and sufficient for guaranteeing liveness. We prove the necessity by a constructive method. Because of the overwhelming complexity of proving the sufficiency of such condition, we assess the correctness of the conjecture in probability through the analysis of the results of a large number of randomized experiments

In this paper we propose a new approach to motion planning, based on the introduction of a lattice structure in the workspace of the robot, leading to efficient computations of plans for rather complex vehicles, and allowing for the implementation of optimization procedures in a rather straightforward way. The basic idea is the purposeful restriction of the set of possible input functions to the vehicle to a finite set of symbols, or control quanta, which, under suitable conditions, generate a regular lattice of reachable points. Once the lattice is generated and a convenient description computed, standard techniques in integer linear programming can be used to find a plan very efficiently. We also provide a correct and complete algorithm to the problem of finding an optimized plan (with respect e.g. to length minimization) consisting in a sequence of graph searches.

This paper considers the problem of solving conflicts arising among several aircraft that are assumed to move in a shared airspace. Aircraft can not get closer to each other than a given safety distance in order to avoid possible conflicts between different airplanes. For such system of multiple aircraft, we consider the path planning problem among given waypoints avoiding all possible conflicts. In particular we are interested in optimal paths, i.e., we want to minimize the total flight time. We propose two different formulations of the multiaircraft conflict avoidance problem as a mixed-integer linear program: in the first case only velocity changes are admissible maneuvers, in the second one only heading angle changes are allowed. Due to the linear formulation of the two problems, solutions may be obtained quickly with standard optimization software, allowing our approach to be implemented in real time.

We consider the problem of efficiently simulating large interconnected mechanical systems. For applications such as haptic rendering of large, complex virtual environments, dynamic simulation software and hardware is still too slow to afford accurate performance in real-time. In particular, mechanisms with closed kinematic chains require solutions for a set of differential equations with algebraic constraints (DAEs) that are often too complex to be computed in real-time by present-day single-processor machines. On the other hand, the structure of most state-of-the-art algorithms does not easily lend itself to parallelization. In this paper, we propose and experimentally verify a technique for DAE simulation that profitably uses a degree of randomization to achieve efficient parallelization.

We consider optimal resolution of air traffic (AT) conflicts. Aircraft are assumed to cruise within a given altitude layer and are modeled as a kinematic system with constant velocity and curvature bounds. Aircraft cannot get closer to each other than a predefined safety distance. For such a system of multiple aircraft, we consider the problem of planning optimal paths among given waypoints. Necessary conditions for optimality of solutions are derived and used to devise a parametrization of possible trajectories that turns into efficient numerical solutions to the problem. Simulation results for a realistic aircraft conflict scenario are provided. A decentralized implementation of the optimal conflict resolution scheme is introduced that may allow free-flight coordination in a cooperative airspace management scheme. Impact of decentralization on performance and safety is finally discussed with the help of extensive simulations.