In Human-Robot Collaboration, safety mechanisms such as Speed and Separation Monitoring and Power and Force Limitation dynamically adjust the robot’s speed based on human proximity. While essential for risk reduction, these mechanisms introduce slowdowns that makes cycle time estimation a hard task and impact job scheduling efficiency. Existing methods for estimating cycle times or designing schedulers often rely on predefined safety models, which may not accurately reflect real-world safety implementations, as these depend on case-specific risk assessments. In this paper, we propose a deep learning approach to predict the robot’s safety scaling factor directly from process execution data. We analyze multiple neural network architectures and demonstrate that a simple feed-forward network effectively estimates the robot’s slowdown. This capability is crucial for improving cycle time predictions and designing more effective scheduling algorithms in collaborative robotic environments.

This paper addresses the optimization of human-robot collaborative work-cells before their physical deployment. Most of the times, such environments are designed based on the experience of the system integrators, often leading to sub-optimal solutions. Accurate simulators of the robotic cell, accounting for the presence of the human as well, are available today and can be used in the pre-deployment. We propose an iterative optimization scheme where a digital model of the work-cell is updated based on a genetic algorithm. The methodology focuses on the layout optimization and task allocation, encoding both the problems simultaneously in the design variables handled by the genetic algorithm, while the task scheduling problem depends on the result of the upper-level one. The final solution balances conflicting objectives in the fitness function and is validated to show the impact of the objectives with respect to a baseline, which represents possible initial choices selected based on the human judgment.

Physics-based simulations and learning-based models are vital for complex robotics tasks like deformable object manipulation and liquid handling. However, these models often struggle with accuracy due to epistemic uncertainty or the sim-to-real gap. For instance, accurately pouring liquid from one container to another poses challenges, particularly when models are trained on limited demonstrations and may perform poorly in novel situations. This paper proposes an uncertainty-aware Monte Carlo Tree Search (MCTS) algorithm designed to mitigate these inaccuracies. By incorporating estimates of model uncertainty, the proposed MCTS strategy biases the search to-wards actions with lower predicted uncertainty. This approach enhances the reliability of planning under uncertain conditions. Applied to a liquid pouring task, our method demonstrates improved success rates even with models trained on minimal data, outperforming traditional methods and showcasing its potential for robust decision-making in robotics.

This paper addresses motion replanning in human-robot collaborative scenarios, with an emphasis on reactivity and safety-compliant efficiency. While existing human-aware motion planners perform well in structured environments, they often struggle with unpredictable human behavior. This can result in safety measures that hinder the robot’s performance and overall throughput. This study combines reactive path replanning and a safety-aware cost function, enabling the robot to adapt its path to the changes in the scene in real-time. This solution reduces the execution time and trajectory slowdowns while ensuring safety. Simulations and real-world experiments show the method’s effectiveness compared to standard human-robot cooperation approaches, with efficiency enhancements of up to 60%.

The collaboration between humans and robots re-quires a paradigm shift not only in robot perception, reasoning, and action, but also in the design of the robotic cell. This paper proposes an optimization framework for designing collaborative robotics cells using a digital twin during the pre-deployment phase. This approach mitigates the limitations of experience-based sub-optimal designs by means of Bayesian optimization to find the optimal layout after a certain number of iterations. By integrating production KPIs into a black-box optimization frame-work, the digital twin supports data-driven decision-making, reduces the need for costly prototypes, and ensures continuous improvement thanks to the learning nature of the algorithm. The paper presents a case study with preliminary results that show how this methodology can be applied to obtain safer, more efficient, and adaptable human-robot collaborative environments.

Robotic manipulation relies on analytical or learned models to simulate the system dynamics. These models are often inaccurate and based on offline information, so that the robot planner is unable to cope with mismatches between the expected and the actual behavior of the system (e.g., the presence of an unexpected obstacle). In these situations, the robot should use information gathered online to correct its planning strategy and adapt to the actual system response. We propose a sampling-based motion planning approach that uses an estimate of the model error and online observations to correct the planning strategy at each new replanning. Our approach adapts the cost function and the sampling bias of a kinodynamic motion planner when the outcome of the executed transitions is different from the expected one (e.g., when the robot unexpectedly collides with an obstacle) so that future trajectories will avoid unreliable motions. To infer the properties of a new transition, we introduce the notion of context-awareness, i.e., we store local environment information for each executed transition and avoid new transitions with context similar to previous unreliable ones. This is helpful for leveraging online information even if the simulated transitions are far (in the state-and-action space) from the executed ones. Simulation and experimental results show that the proposed approach increases the success rate in execution and reduces the number of replannings needed to reach the goal.

Combining symbolic and geometric reasoning in multiagent systems is a challenging task that involves planning, scheduling, and synchronization problems. Existing works overlooked the variability of task duration and geometric feasibility intrinsic to these systems because of the interaction between agents and the environment. We propose a combined task and motion planning approach to optimize the sequencing, assignment, and execution of tasks under temporal and spatial variability. The framework relies on decoupling tasks and actions, where an action is one possible geometric realization of a symbolic task. At the task level, timeline-based planning deals with temporal constraints, duration variability, and synergic assignment of tasks. At the action level, online motion planning plans for the actual movements dealing with environmental changes. We demonstrate the approach's effectiveness in a collaborative manufacturing scenario, in which a robotic arm and a human worker shall assemble a mosaic in the shortest time possible. Compared with existing works, our approach applies to a broader range of applications and reduces the execution time of the process.

With the spread of robots in unstructured, dynamic environments, the topic of path replanning has gained importance in the robotics community. Although the number of replanning strategies has significantly increased, there is a lack of agreed-upon libraries and tools, making the use, development, and benchmarking of new algorithms arduous. This paper introduces OpenMORE, a new open-source ROS-based C++ library for sampling-based path replanning algorithms. The library builds a framework that allows for continuous replanning and collision checking of the traversed path during the execution of the robot trajectory. Users can solve replanning tasks exploiting the already available algorithms and can easily integrate new ones, leveraging the library to manage the entire execution.

This paper addresses human-robot collaboration (HRC) challenges of integrating predictions of human activity to provide a proactive-n-reactive response capability for the robot. Prior works that consider current or predicted human poses as static obstacles are too nearsighted or too conservative in planning, potentially causing delayed robot paths. Alternatively, time-varying prediction of human poses would enable robot paths that avoid anticipated human poses, synchronized dynamically in time and space. Herein, a proactive path planning method, denoted STAP, is presented that uses spatiotemporal human occupancy maps to find robot trajectories that anticipate human movements, allowing robot passage without stopping. In addition, STAP anticipates delays from robot speed restrictions required by ISO/TS 15066 speed and separation monitoring (SSM). STAP also proposes a sampling-based planning algorithm based on RRT* to solve the spatio-temporal motion planning problem and find paths of minimum expected duration. Experimental results show STAP generates paths of shorter duration and greater average robot-human separation distance throughout tasks. Additionally, STAP more accurately estimates robot trajectory durations in HRC, which are useful in arriving at proactive-n-reactive robot sequencing.

A good estimation of the actions’ cost is key in task planning for human-robot collaboration. The duration of an action depends on agents’ capabilities and the correlation between actions performed simultaneously by the human and the robot. This paper proposes an approach to learning actions’ costs and coupling between actions executed concurrently by humans and robots. We leverage the information from past executions to learn the average duration of each action and a synergy coefficient representing the effect of an action performed by the human on the duration of the action performed by the robot (and vice versa). We implement the proposed method in a simulated scenario where both agents can access the same area simultaneously. Safety measures require the robot to slow down when the human is close, denoting a bad synergy of tasks operating in the same area. We show that our approach can learn such bad couplings so that a task planner can leverage this information to find better plans.

Robots working in proximity of humans often need to change their motion to avoid collisions and interference with the operators. This paper uses a path re-planning approach to change the robot path online when the human operator is in the robot way. The method exploits a set of pre-computed paths to compute a new feasible path in case of obstruction to enhance the trajectory’s readability. Moreover, the algorithm iteratively optimizes the current solution in an anytime fashion to deal with strict computing time requirements. Experimental results show the method’s effectiveness in a collaborative cell, compared with industry best practices.

Combining task and motion planning efficiently in human-robot collaboration (HRC) entails several challenges because of the uncertainty conveyed by the human behavior. Tasks plan execution should be continuously monitored and updated based on the actual behavior of the human and the robot to maintain productivity and safety. We propose control-based approach based on two layers, i.e., task planning and action planning. Each layer reasons at a different level of abstraction: task planning considers high-level operations without taking into account their motion properties; action planning optimizes the execution of high-level operations based on current human state and geometric reasoning. The result is a hierarchical framework where the bottom layer gives feedback to top layer about the feasibility of each task, and the top layer uses this feedback to (re)optimize the process plan. The method is applied to an industrial case study in which a robot and a human worker cooperate to assemble a mosaic.

Human robot collaboration requires new planning strategies to guarantee an efficient and safe coexistence of robots and humans in the workspace. We propose a framework based on a model predictive control approach to trajectory scaling and inverse kinematics. The online modification of the velocity override slows down the task to ensure safety and the redundancy of the system is exploited to maximize the distance from the operator. Experimental results on a 7-degree-of-freedom robotic system prove the effectiveness of the method.

The paper presents a fast online predictive method to solve the task-priority differential inverse kinematics of redundant manipulators under kinematic constraints. It implements a task-scaling technique to preserve the desired geometrical task, when the trajectory is infeasible for the robot capabilities. Simulation results demonstrate the effectiveness of the methodology.

Reduction of the energy consumption in robotized processes is a key issue in nowadays manufacturing. In this paper, we propose a simple approach to energy minimization of robotic tasks with assigned cycle time based on sequential quadratic programming. The method aims at re-shaping a given timing law in the sense of energy saving, without modifying the desired path and the given cycle time. Thanks to the iterative linearization of the nonlinear time-constraint, the resulting minimization problem is solved by only using common quadratic programming solvers, making the method suitable for a direct implementation in robot industrial controllers. At first, the method is devised by only considering the kinematics of the manipulator. The dynamic model is then straightforwardly included, without significantly increasing the complexity of the method. Validation in simulation environment is provided in order to show the effectiveness of the methodology.

In this paper, a Model Predictive Control approach for the velocity control of operator-in-the loop overhead cranes is proposed. The operator can select the maximum position overshoot as a tuning parameter for the method. Simulations provide a comparison between the proposed method and the well known Zero Vibration input shaping technique, showing its effectiveness in controlling the payload oscillations.

In this paper we present a new method to reduce the computational complexity of model predictive control algorithms with online optimization. The formulation of the predictive equations is performed in the continuous-time domain, while the control inputs are parametrized as piecewise constant functions, with less steps than the control horizon. The continuous-time formulation permits the arbitrary choice of the prediction and control time-instants, disregarding the sampling period of the system, and this improves the goodness of the approximation. Moreover, since the inputs are forced to be piecewise constant, the resulting controller can be directly implemented in discretetime. A tuning method for the choice of the prediction and control instants is proposed, minimizing the deviation with respect to the non-approximated controller. Numerical results show the effectiveness of this strategy against other methods with same reduction of computational complexity.

In this paper, we present a new approach to manipulability maximisation for a dual-arm manipulator, which takes into account the manipulability of the overall task. This method tries to overcome the drawbacks given by traditional approaches, which optimise the manipulability of the local configuration of the manipulator, but do not take into account the rest of the task, even though it is known a priori. In this way, it is possible to improve the average manipulability index over the task. The method is applied to a dual-arm system, wherein the task is expressed in terms of relative poses between the end-effectors. For this reason, the kinematic of the system is solved by means of the relative Jacobian.