-
Risk-Based Safety Envelopes for Autonomous Vehicles Under Perception Uncertainty
Authors:
Julian Bernhard,
Patrick Hart,
Amit Sahu,
Christoph Schöller,
Michell Guzman Cancimance
Abstract:
Ensuring the safety of autonomous vehicles, given the uncertainty in sensing other road users, is an open problem. Moreover, separate safety specifications for perception and planning components raise how to assess the overall system safety. This work provides a probabilistic approach to calculate safety envelopes under perception uncertainty. The probabilistic envelope definition is based on a ri…
▽ More
Ensuring the safety of autonomous vehicles, given the uncertainty in sensing other road users, is an open problem. Moreover, separate safety specifications for perception and planning components raise how to assess the overall system safety. This work provides a probabilistic approach to calculate safety envelopes under perception uncertainty. The probabilistic envelope definition is based on a risk threshold. It limits the cumulative probability that the actual safety envelope in a fully observable environment is larger than an applied envelope and is solved using iterative worst-case analysis of envelopes. Our approach extends non-probabilistic envelopes - in this work, the Responsibility-Sensitive Safety (RSS) - to handle uncertainties. To evaluate our probabilistic envelope approach, we compare it in a simulated highway merging scenario against several baseline safety architectures. Our evaluation shows that our model allows adjusting safety and performance based on a chosen risk level and the amount of perception uncertainty. We conclude with an outline of how to formally argue safety under perception uncertainty using our formulation of envelope violation risk.
△ Less
Submitted 21 July, 2021;
originally announced July 2021.
-
FloMo: Tractable Motion Prediction with Normalizing Flows
Authors:
Christoph Schöller,
Alois Knoll
Abstract:
The future motion of traffic participants is inherently uncertain. To plan safely, therefore, an autonomous agent must take into account multiple possible trajectory outcomes and prioritize them. Recently, this problem has been addressed with generative neural networks. However, most generative models either do not learn the true underlying trajectory distribution reliably, or do not allow predict…
▽ More
The future motion of traffic participants is inherently uncertain. To plan safely, therefore, an autonomous agent must take into account multiple possible trajectory outcomes and prioritize them. Recently, this problem has been addressed with generative neural networks. However, most generative models either do not learn the true underlying trajectory distribution reliably, or do not allow predictions to be associated with likelihoods. In our work, we model motion prediction directly as a density estimation problem with a normalizing flow between a noise distribution and the future motion distribution. Our model, named FloMo, allows likelihoods to be computed in a single network pass and can be trained directly with maximum likelihood estimation. Furthermore, we propose a method to stabilize training flows on trajectory datasets and a new data augmentation transformation that improves the performance and generalization of our model. Our method achieves state-of-the-art performance on three popular prediction datasets, with a significant gap to most competing models.
△ Less
Submitted 30 July, 2021; v1 submitted 5 March, 2021;
originally announced March 2021.
-
Providentia -- A Large-Scale Sensor System for the Assistance of Autonomous Vehicles and Its Evaluation
Authors:
Annkathrin Krämmer,
Christoph Schöller,
Dhiraj Gulati,
Venkatnarayanan Lakshminarasimhan,
Franz Kurz,
Dominik Rosenbaum,
Claus Lenz,
Alois Knoll
Abstract:
The environmental perception of an autonomous vehicle is limited by its physical sensor ranges and algorithmic performance, as well as by occlusions that degrade its understanding of an ongoing traffic situation. This not only poses a significant threat to safety and limits driving speeds, but it can also lead to inconvenient maneuvers. Intelligent Infrastructure Systems can help to alleviate thes…
▽ More
The environmental perception of an autonomous vehicle is limited by its physical sensor ranges and algorithmic performance, as well as by occlusions that degrade its understanding of an ongoing traffic situation. This not only poses a significant threat to safety and limits driving speeds, but it can also lead to inconvenient maneuvers. Intelligent Infrastructure Systems can help to alleviate these problems. An Intelligent Infrastructure System can fill in the gaps in a vehicle's perception and extend its field of view by providing additional detailed information about its surroundings, in the form of a digital model of the current traffic situation, i.e. a digital twin. However, detailed descriptions of such systems and working prototypes demonstrating their feasibility are scarce. In this paper, we propose a hardware and software architecture that enables such a reliable Intelligent Infrastructure System to be built. We have implemented this system in the real world and demonstrate its ability to create an accurate digital twin of an extended highway stretch, thus enhancing an autonomous vehicle's perception beyond the limits of its on-board sensors. Furthermore, we evaluate the accuracy and reliability of the digital twin by using aerial images and earth observation methods for generating ground truth data.
△ Less
Submitted 8 December, 2021; v1 submitted 16 June, 2019;
originally announced June 2019.
-
Targetless Rotational Auto-Calibration of Radar and Camera for Intelligent Transportation Systems
Authors:
Christoph Schöller,
Maximilian Schnettler,
Annkathrin Krämmer,
Gereon Hinz,
Maida Bakovic,
Müge Güzet,
Alois Knoll
Abstract:
Most intelligent transportation systems use a combination of radar sensors and cameras for robust vehicle perception. The calibration of these heterogeneous sensor types in an automatic fashion during system operation is challenging due to differing physical measurement principles and the high sparsity of traffic radars. We propose - to the best of our knowledge - the first data-driven method for…
▽ More
Most intelligent transportation systems use a combination of radar sensors and cameras for robust vehicle perception. The calibration of these heterogeneous sensor types in an automatic fashion during system operation is challenging due to differing physical measurement principles and the high sparsity of traffic radars. We propose - to the best of our knowledge - the first data-driven method for automatic rotational radar-camera calibration without dedicated calibration targets. Our approach is based on a coarse and a fine convolutional neural network. We employ a boosting-inspired training algorithm, where we train the fine network on the residual error of the coarse network. Due to the unavailability of public datasets combining radar and camera measurements, we recorded our own real-world data. We demonstrate that our method is able to reach precise and robust sensor registration and show its generalization capabilities to different sensor alignments and perspectives.
△ Less
Submitted 28 July, 2019; v1 submitted 18 April, 2019;
originally announced April 2019.
-
What the Constant Velocity Model Can Teach Us About Pedestrian Motion Prediction
Authors:
Christoph Schöller,
Vincent Aravantinos,
Florian Lay,
Alois Knoll
Abstract:
Pedestrian motion prediction is a fundamental task for autonomous robots and vehicles to operate safely. In recent years many complex approaches based on neural networks have been proposed to address this problem. In this work we show that - surprisingly - a simple Constant Velocity Model can outperform even state-of-the-art neural models. This indicates that either neural networks are not able to…
▽ More
Pedestrian motion prediction is a fundamental task for autonomous robots and vehicles to operate safely. In recent years many complex approaches based on neural networks have been proposed to address this problem. In this work we show that - surprisingly - a simple Constant Velocity Model can outperform even state-of-the-art neural models. This indicates that either neural networks are not able to make use of the additional information they are provided with, or that this information is not as relevant as commonly believed. Therefore, we analyze how neural networks process their input and how it impacts their predictions. Our analysis reveals pitfalls in training neural networks for pedestrian motion prediction and clarifies false assumptions about the problem itself. In particular, neural networks implicitly learn environmental priors that negatively impact their generalization capability, the motion history of pedestrians is irrelevant and interactions are too complex to predict. Our work shows how neural networks for pedestrian motion prediction can be thoroughly evaluated and our results indicate which research directions for neural motion prediction are promising in future.
△ Less
Submitted 22 January, 2020; v1 submitted 19 March, 2019;
originally announced March 2019.