In this work, we address the problem of formal safety verification for stochastic cyber-physical systems (CPS) equipped with ReLU neural network (NN) controllers. Our goal is to find the set of initial states from where, with a predetermined confidence, the system will not reach an unsafe configuration within a specified time horizon. Specifically, we consider discrete-time LTI systems with Gaussian noise, which we abstract by a suitable graph. Then, we formulate a Satisfiability Modulo Convex (SMC) problem to estimate upper bounds on the transition probabilities between nodes in the graph. Using this abstraction, we propose a method to compute tight bounds on the safety probabilities of nodes in this graph, despite possible over-approximations of the transition probabilities between these nodes. Additionally, using the proposed SMC formula, we devise a heuristic method to refine the abstraction of the system in order to further improve the estimated safety bounds. Finally, we corroborate the efficacy of the proposed method with simulation results considering a robot navigation example and comparison against a state-of-the-art verification scheme.

In this work, we present a decentralized motion planning and control architecture for the cooperative loading task using heterogeneous robotic agents operating in a cluttered workspace with static obstacles. Initially, we tackle the problem of calculating a set of feasible loading configurations via a Probabilistic Road Maps technique. Next, an optimal loading configuration is selected considering the connectivity of the space and the Euclidean distance between the robotic agents. A motion control scheme for each agent is designed and implemented in order to autonomously guide each robot to the desired loading configuration with guaranteed obstacle avoidance and convergence properties. The performance and the applicability of the proposed strategy is experimentally verified in a variety of loading scenarios using a redundant static manipulator and a mobile platform.

In this work, a hybrid control scheme is presented in order to address the navigation problem for a planar robotic platform of arbitrary shape that is moving inside an obstacle cluttered workspace. Given an initial and desired robot configuration, we propose a methodology based on approximate configuration space decomposition techniques that makes use of heuristics to adaptively refine a partition of the configuration space into non-overlapping, adjacent slices. Furthermore, we employ appropriate workspace transformations and adaptive potential field based control laws that integrate elegantly with the type of configuration space representation used, in order to safely navigate within a given cell and successfully cross over to the next, for almost all initial configurations, until the desired configuration is reached. Finally, we present simulation results that demonstrate the efficacy of the proposed control scheme.

Artificial Potential Fields (APFs) constitute an intuitive tool for designing autonomous robot navigation control schemes, though they generally suffer from the existence of local minima which may trap the robot away from its desired configuration, an issue usually addressed by appropriate offline “tuning” of the potential field's parameters. On the other side, most APF based approaches rely on a diffeomorphism to sphere worlds to handle realistic scenarios, which may be either costly to compute (e.g., conformal mappings) or requires some sort of preconditioning of the workspace (e.g., decomposition of complex geometries to simple elementary components). In this work, we first propose a constructive procedure to map multiply connected compact 2D workspaces to one or more punctured disks based on harmonic maps. Subsequently, we design an APF based control scheme along with an adaptive law for its parameters that requires no offline tuning to guarantee safe convergence to its goal configuration. Finally, an extensive simulation study is conducted to demonstrate the efficacy of the proposed control scheme.

This paper presents the design of a vision-based object grasping and motion control architecture for a mobile manipulator system. The optimal grasping areas of the object are estimated using the partial point cloud acquired from an onboard RGB-D sensor system. The reach-to-grasp motion of the mobile manipulator is handled via a Nonlinear Model Predictive Control scheme. The controller is formulated accordingly in order to allow the system to operate in a constrained workspace with static obstacles. The goal of the proposed scheme is to guide the robot's end-effector towards the optimal grasping regions with guaranteed input and state constraints such as occlusion and obstacle avoidance, workspace boundaries and field of view constraints. The performance of the proposed strategy is experimentally verified using an 8 Degrees of Freedom KUKA Youbot in different reach-to-grasp scenarios.

This paper addresses the fault tolerant control problem for an omni-directional mobile platform with four mecanum wheels moving on a well-known flat and constrained workspace with static obstacles. As a fault, we consider the case where a wheel cannot be actuated and hence it rotates freely around its drive shaft owing to the friction with the flat surface. Depending on the multitude of the faults, a robust motion control scheme is developed that achieves any desired configuration within the operational workspace, avoids collisions with the obstacles and does not violate the workspace boundaries despite the presence of dynamic model uncertainties. The challenge with respect to the current state of the art in fault tolerant control for such mobile platforms, where only one faulty wheel has been considered (i.e., the platform still retains its full actuation capabilities), lies in completely compensating up to two faulty wheels (i.e., the model becomes underactuated in this way) despite the dynamic model uncertainty and the presence of static obstacles in the workspace. Navigation Functions are innovatively incorporated with adaptive control techniques to deal with the parametric uncertainty in the robot dynamics, extending thus greatly the current state of the art in robust motion planning and collision avoidance by studying second order dynamics with parametric uncertainty. Finally, an extensive experimental study clarifies the proposed method and verifies its efficiency in various faults.

In this work we study the problem of landing a quadrotor on an inclined moving platform. The aerial robot employs an forward looking on-board camera to detect and observe the landing platform, which is carried by a mobile robot moving independently on an inclined surface. The platform may also be tilted with respect to the mobile robot. The overall goal is to design the aerial robot's control inputs such that it initially approaches the platform, while maintaining it within the camera's field of view and finally lands on it, in a way that minimizes the errors in position, attitude and velocity, while avoiding collision. Owing to the inclined ground and landing surface, the desired final state of the aerial robot is not an equilibrium state, which complicates significantly the control design. In that respect, a discrete-time non-linear model predictive controller was developed that optimizes both the trajectories and the time horizon, towards achieving the aforementioned objectives while respecting the input constraints as well. Finally, an extensive experimental study, with a Pioneer mobile robot and a Parrot ARDrone quadrotor, clarifies and verifies the theoretical findings.