arXiv subjects

Laura Ferranti

Publications and source records attributed to Laura Ferranti.

27 records · Page 2Linked to original sources

arXiv2023

Optimization-based Fault Mitigation for Safe Automated Driving

With increased developments and interest in cooperative driving and higher levels of automation (SAE level 3+), the need for safety systems that are capable to monitor system health and maintain safe operations in faulty scenarios is increasing. A variety of faults or failures could occur, and there exists a high variety of ways to respond to such events. Once a fault or failure is detected, there is a need to classify its severity and decide on appropriate and safe mitigating actions. To provide a solution to this mitigation challenge, in this paper a functional-safety architecture is proposed and an optimization-based mitigation algorithm is introduced. This algorithm uses nonlinear model predictive control (NMPC) to bring a vehicle, suffering from a severe fault, such as a power steering failure, to a safe-state. The internal model of the NMPC uses the information from the fault detection, isolation and identification to optimize the tracking performance of the controller, showcasing the need of the proposed architecture. Given a string of ACC vehicles, our results demonstrate a variety of tactical decision-making approaches that a fault-affected vehicle could employ to manage any faults. Furthermore, we show the potential for improving the safety of the affected vehicle as well as the effect of these approaches on the duration of the manoeuvre.

math.OC↗

arXiv2022

PARTNR: Pick and place Ambiguity Resolving by Trustworthy iNteractive leaRning

Several recent works show impressive results in mapping language-based human commands and image scene observations to direct robot executable policies (e.g., pick and place poses). However, these approaches do not consider the uncertainty of the trained policy and simply always execute actions suggested by the current policy as the most probable ones. This makes them vulnerable to domain shift and inefficient in the number of required demonstrations. We extend previous works and present the PARTNR algorithm that can detect ambiguities in the trained policy by analyzing multiple modalities in the pick and place poses using topological analysis. PARTNR employs an adaptive, sensitivity-based, gating function that decides if additional user demonstrations are required. User demonstrations are aggregated to the dataset and used for subsequent training. In this way, the policy can adapt promptly to domain shift and it can minimize the number of required demonstrations for a well-trained policy. The adaptive threshold enables to achieve the user-acceptable level of ambiguity to execute the policy autonomously and in turn, increase the trustworthiness of our system. We demonstrate the performance of PARTNR in a table-top pick and place task.

cs.RO↗

arXiv2022

Learning Mixed Strategies in Trajectory Games

In multi-agent settings, game theory is a natural framework for describing the strategic interactions of agents whose objectives depend upon one another's behavior. Trajectory games capture these complex effects by design. In competitive settings, this makes them a more faithful interaction model than traditional "predict then plan" approaches. However, current game-theoretic planning methods have important limitations. In this work, we propose two main contributions. First, we introduce an offline training phase which reduces the online computational burden of solving trajectory games. Second, we formulate a lifted game which allows players to optimize multiple candidate trajectories in unison and thereby construct more competitive "mixed" strategies. We validate our approach on a number of experiments using the pursuit-evasion game "tag."

cs.GT↗

arXiv2022

Where to Look Next: Learning Viewpoint Recommendations for Informative Trajectory Planning

Search missions require motion planning and navigation methods for information gathering that continuously replan based on new observations of the robot's surroundings. Current methods for information gathering, such as Monte Carlo Tree Search, are capable of reasoning over long horizons, but they are computationally expensive. An alternative for fast online execution is to train, offline, an information gathering policy, which indirectly reasons about the information value of new observations. However, these policies lack safety guarantees and do not account for the robot dynamics. To overcome these limitations we train an information-aware policy via deep reinforcement learning, that guides a receding-horizon trajectory optimization planner. In particular, the policy continuously recommends a reference viewpoint to the local planner, such that the resulting dynamically feasible and collision-free trajectories lead to observations that maximize the information gain and reduce the uncertainty about the environment. In simulation tests in previously unseen environments, our method consistently outperforms greedy next-best-view policies and achieves competitive performance compared to Monte Carlo Tree Search, in terms of information gains and coverage time, with a reduction in execution time by three orders of magnitude.

cs.RO↗

arXiv2021

DeepKoCo: Efficient latent planning with a task-relevant Koopman representation

This paper presents DeepKoCo, a novel model-based agent that learns a latent Koopman representation from images. This representation allows DeepKoCo to plan efficiently using linear control methods, such as linear model predictive control. Compared to traditional agents, DeepKoCo learns task-relevant dynamics, thanks to the use of a tailored lossy autoencoder network that allows DeepKoCo to learn latent dynamics that reconstruct and predict only observed costs, rather than all observed dynamics. As our results show, DeepKoCo achieves similar final performance as traditional model-free methods on complex control tasks while being considerably more robust to distractor dynamics, making the proposed agent more amenable for real-life applications.

cs.LG↗

arXiv2021

Active Safety System for Semi-Autonomous Teleoperated Vehicles

Autonomous cars can reduce road traffic accidents and provide a safer mode of transport. However, key technical challenges, such as safe navigation in complex urban environments, need to be addressed before deploying these vehicles on the market. Teleoperation can help smooth the transition from human operated to fully autonomous vehicles since it still has human in the loop providing the scope of fallback on driver. This paper presents an Active Safety System (ASS) approach for teleoperated driving. The proposed approach helps the operator ensure the safety of the vehicle in complex environments, that is, avoid collisions with static or dynamic obstacles. Our ASS relies on a model predictive control (MPC) formulation to control both the lateral and longitudinal dynamics of the vehicle. By exploiting the ability of the MPC framework to deal with constraints, our ASS restricts the controller's authority to intervene for lateral correction of the human operator's commands, avoiding counter-intuitive driving experience for the human operator. Further, we design a visual feedback to enhance the operator's trust over the ASS. In addition, we propose an MPC's prediction horizon data based novel predictive display to mitigate the effects of large latency in the teleoperation system. We tested the performance of the proposed approach on a high-fidelity vehicle simulator in the presence of dynamic obstacles and latency.

cs.RO↗

arXiv2020

Model Predictive Contouring Control for Collision Avoidance in Unstructured Dynamic Environments

This paper presents a method for local motion planning in unstructured environments with static and moving obstacles, such as humans. Given a reference path and speed, our optimization-based receding-horizon approach computes a local trajectory that minimizes the tracking error while avoiding obstacles. We build on nonlinear model-predictive contouring control (MPCC) and extend it to incorporate a static map by computing, online, a set of convex regions in free space. We model moving obstacles as ellipsoids and provide a correct bound to approximate the collision region, given by the Minkowsky sum of an ellipse and a circle. Our framework is agnostic to the robot model. We present experimental results with a mobile robot navigating in indoor environments populated with humans. Our method is executed fully onboard without the need of external support and can be applied to other robot morphologies such as autonomous cars.

cs.RO↗

arXiv2016

Asynchronous Splitting Design for Model Predictive Control

This paper focuses on the design of an asynchronous dual solver suitable for embedded model predictive control (MPC) applications. The proposed solver relies on a state-of-the-art variance reduction (VR) scheme, previously used in the context of stochastic proximal gradient methods, and on the alternating minimization algorithm (AMA). The resultant algorithm, a stochastic AMA with VR, shows geometric convergence (in the expectation) to a suboptimal solution of the MPC problem and, compared to other state-of-the-art dual asynchronous algorithms, allows to tune the probability of the asynchronous updates to improve the quality of the estimates. We apply the proposed algorithm to a specific class of splitting methods, i.e., the decomposition along the length of the prediction horizon, and provide preliminary numerical results on a practical application, the longitudinal control of an Airbus passenger aircraft.

math.OC↗

arXiv2015

A Parallel Dual Fast Gradient Method for MPC Applications

We propose a parallel adaptive constraint-tightening approach to solve a linear model predictive control problem for discrete-time systems, based on inexact numerical optimization algorithms and operator splitting methods. The underlying algorithm first splits the original problem in as many independent subproblems as the length of the prediction horizon. Then, our algorithm computes a solution for these subproblems in parallel by exploiting auxiliary tightened subproblems in order to certify the control law in terms of suboptimality and recursive feasibility, along with closed-loop stability of the controlled system. Compared to prior approaches based on constraint tightening, our algorithm computes the tightening parameter for each subproblem to handle the propagation of errors introduced by the parallelization of the original problem. Our simulations show the computational benefits of the parallelization with positive impacts on performance and numerical conditioning when compared with a recent nonparallel adaptive tightening scheme.

math.OC↗