In a wide range of domains, such as pipeline inspection, surveillance in smart cities and tracking of multiple microparticles by an optical microscope, a common goal is to use mobile agents to persistently monitor a set of targets. We refer to this as the persistent monitoring problem. In this dissertation, we assume that each of these targets has an internal state that evolves with linear stochastic dynamics. The agents can observe these states when they are close to the targets, and the goal is to plan agent trajectories such that the sensed data can be used to minimize the uncertainty of the estimation process. We study scalable approaches for planning agent trajectories that minimize the long term uncertainty of the target states. We design algorithms that are computationally efficient and simple to implement, but grounded in mathematically proven performance guarantees. First we approach the problem from a continuous time perspective with the goal of finding locally optimal agent trajectories using a gradient descent scheme. We assume that trajectories are fully defined by a finite set of parameters and compute the cost gradients. Considering periodic agent trajectories and an infinite time horizon, we prove that, under some natural assumptions, the uncertainty of each target converges to a limit cycle. We also show that, in 1D environments with bounded controls, an optimal control is parametric. In multidimensional settings, we propose an efficient parameterization using Fourier curves. Simulation results show the efficiency of our approach. Next, we consider a graph-constrained, single-agent version of the problem, where agents can only move in the edges of the graph and observe the target when they are visiting the node corresponding to it. We prove that, in this scenario, an optimal policy is such that all the agent have a common peak uncertainty. Using this property of the optimal solution, we develop lightweight algorithms that, instead of directly solving the optimization problem, balance the dwelling times to fulfill such property of an optimal policy. In some particular situations, global optimality of the proposed algorithm is proven. Using a custom-designed greedy exploration scheme, we develop an efficient method for obtaining efficient target visiting sequences. We extended this approach to multi-agent scenarios by using a divide-and conquer strategy, where targets are divided in clusters and each of these clusters is only visited by one agent. Then, we extend those ideas to a discrete time version of the problem. We show that, for a periodic trajectory with fixed cycle length, the problem can be formulated as set of semidefinite programs. This allowed us to leverage efficient SDP solvers to provide fast solutions to the persistent monitoring problem. We design a scheme that leverages the spatial configuration of the targets to guide the search over this set of optimization problems to provide efficient trajectories. Finally we describe an application of the proposed techniques to the problem of tracking multiple diffusing particles using a feedback-driven confocal microscope. The proposed persistent monitoring algorithm was used as the higher level controller in a hierarchical scheme, defining which particle should be tracked at each instant. Then an extremum seeking controller was used as a lower level controller in order to track the moving particle and provide efficient observations.

No description available.

An autonomous robot system requires robots to actively gather information using sensors in order to make control decisions. Some problems where autonomous robots are useful include mapping, environmental monitoring, and surveillance. In some cases, information gathering turns into a multiple target tracking (MTT) problem. Usually, an MTT tracker is utilized to recursively estimate both the number of targets and the state of each target. In order to estimate more efficiently and reliably, sensors must balance exploiting current knowledge to track known targets while simultaneously exploring to find information about new targets. This yields to the coverage control problem, which is aimed at maximizing the total sensing capability of a sensing network over the entire mission space. Many applications of sensing networks benefit from utilizing distributed manners, in which cases networks are able to be scaled to large swarms and better tolerate failures of individual sensors. A distributed network requires sensors to exchange data locally and cooperate in decision making globally.This dissertation studies MTT based on random finite set (RFS) for iterative target states estimation and Voronoi-based coverage control algorithms for target tracking. We address a series of four main problems aiming at allowing reliable and efficient target tracking for distributed multi-robot systems in complicated real-world scenarios and push forward the realization of robot coordination techniques. Firstly, we propose novel target estimation and coverage control schemes to incorporate robots with localization uncertainty. Secondly, we improve target search efficiency for teams of robot with no prior knowledge of target models or distributions by enabling active search and environment learning. Thirdly, we allow robots with heterogeneous capacities in perception and kinematics to cooperatively search and track in an efficient way. Lastly, we develop an improved MTT tracker to allow estimating semantic object labels over time. The efficacy of the proposed methods has been validated in series of simulations and/or hardware validations. / Mechanical Engineering

Linear programming is an established, well-understood technique optimization problem; the goal of this thesis is to show that we can still use linear programming to advance the state of the art in two important blocks of modern robotic systems, namely perception, and control. In the context of perception, we study the effects of outliers in the solution of localization problems. In its essence, this problem reduces to finding the coordinates of a set of nodes in a common reference frame starting from relative pairwise measurements and is at the core of many applications such as Structure from Motion (SfM), sensor networks, and Simultaneous Localization And Mapping (SLAM). In practical situations, the accuracy of the relative measurements is marred by noise and outliers (large-magnitude errors). In particular, outliers might introduce significant errors in the final result, hence, we have the problem of quantifying how much we should trust the solution returned by some given localization solver. In this work, we focus on the question of whether an L1-norm robust optimization formulation can recover a solution that is identical to the ground truth, under the scenario of translation-only measurements corrupted exclusively by outliers and no noise. In the context of control, we study the problem of robust path planning. Path planning deals with the problem of finding a path from an initial state toward a goal state while considering collision avoidance. We propose a novel approach for navigating in polygonal environments by synthesizing controllers that take as input relative displacement measurements with respect to a set of landmarks. Our algorithm is based on solving a sequence of robust min-max Linear Programming problems on the elements of a cell decomposition of the environment. The optimization problems are formulated using linear Control Lyapunov Function (CLF) and Control Barrier Function (CBF) constraints, to provide stability and safety guarantees, respectively. We integrate the CBF and CLF constraints with sampling-based path planning methods to omit the assumption of having a polygonal environment and add implementation to learn the constraints and estimate the controller when the environment is not fully known. We introduce a method to find the controller synthesis using bearing-only measurements in order to use monocular camera measurements. We show through simulations that the resulting controllers are robust to significant deformations of the environment. These works provide a simple approach in terms of computation to study the robustness of the localization and navigation problem.

Soft optical sensing strategies are rapidly developing for soft robotic systems. In this thesis, a roughness tuning strategy for the fabrication of soft optical sensors to achieve the triple functionality of shape sensing combined with contact detection and force intensity recognition within a single multi-modal sensor is presented. The integration of these sensors into a fully soft robotic platform is demonstrated. The robot consists of a multi-directional bending module with integrated 3D shape sensing and a gripper with tip position monitoring along with contact force recognition and sensing. This robot is tested in a mock laparoscopic setup to prove the effectiveness to be implemented into minimally invasive medical robotic systems. / 2025-01-15T00:00:00Z

Multi-agent systems (MAS) are used as models for many natural and engineered systems, such as robotic teams and cell-cell interactions. Such systems exhibit time-varying spatial (spatio-temporal) behaviors. As the complexity of MAS increases, there is a need to express their behaviors in formal ways that are interpretable to humans and amenable to rigorous mathematical analysis. In this thesis, we propose using spatio-temporal (ST) logics to write up such expressions. In addition, we address two closely related challenges 1) inferring ST logic expressions from data (the inference problem) and 2) synthesizing system inputs such that the MAS outputs meet specific behavioral requirements given by ST logic expressions (the synthesis problem). We consider two distinct MAS types 1) patterning chemical and biological systems and 2) robotic teams. Overall, this thesis has three main parts. First, we develop ST logics that are (1) capable of describing emerging MAS behaviors and (2) equipped with qualitative and quantitative (robustness metric) semantics. The qualitative semantics address the question "are the requirements satisfied/violated?" while the quantitative semantics address the question "how well are the requirements satisfied/violated?" Second, we develop several techniques for inferring ST logics expressions from executions of patterning systems. The proposed techniques utilize unsupervised and supervised learning techniques to learn the structure and parameters of logical expressions. Third, we propose several methods to solve the synthesis problem when requirements are given by the ST logic formulae. We formulate the synthesis problems as optimization problems where the objective is to maximize the robustness metric, thus satisfying the requirements. We outline our approach for solving optimization problems and learning controllers using optimization and deep learning techniques. We demonstrate the efficacy of the proposed algorithms and tools in simulated examples of patterning systems and robotic teams. We conclude with a discussion about the limitations and future research directions. / 2025-01-16T00:00:00Z

No description available.

No description available.

No description available.

No description available.