Simulation and the monte carlo method

With the objective to improve road safety, the automotive industry is moving toward more “intelligent” vehicles. One of the major challenges is to detect dangerous situations and react accordingly in order to avoid or mitigate accidents. This requires predicting the likely evolution of the current traffic situation, and assessing how dangerous that future situation might be. This paper is a survey of existing methods for motion prediction and risk assessment for intelligent vehicles. The proposed classification is based on the semantics used to define motion and risk. We point out the tradeoff between model completeness and real-time constraints, and the fact that the choice of a risk assessment method is influenced by the selected motion model.

/pdf/a-survey-on-motion-prediction-and-risk-assessment-for-39io9df2mm.pdf

A survey on motion prediction and risk assessment for intelligent vehicles

Environment models serve as important resources for an autonomous robot by providing it with the necessary task-relevant information about its habitat. Their use enables robots to perform their tasks more reliably, flexibly, and efficiently. As autonomous robotic platforms get more sophisticated manipulation capabilities, they also need more expressive and comprehensive environment models: for manipulation purposes their models have to include the objects present in the world, together with their position, form, and other aspects, as well as an interpretation of these objects with respect to the robot tasks. The dissertation presented in this article (Rusu, PhD thesis, 2009) proposes Semantic 3D Object Models as a novel representation of the robot’s operating environment that satisfies these requirements and shows how these models can be automatically acquired from dense 3D range data.

/pdf/semantic-3d-object-maps-for-everyday-manipulation-in-human-van9xaq8u7.pdf

Semantic 3D Object Maps for Everyday Manipulation in Human Living Environments

Automatic reconstruction of as-built building information models from laser-scanned point clouds: A review of related techniques

Building information models (BIMs) are maturing as a new paradigm for storing and exchanging knowledge about a facility. BIMs constructed from a CAD model do not generally capture details of a facility as it was actually built. Laser scanners can be used to capture dense 3D measurements of a facility's as-built condition and the resulting point cloud can be manually processed to create an as-built BIM — a time-consuming, subjective, and error-prone process that could benefit significantly from automation. This article surveys techniques developed in civil engineering and computer science that can be utilized to automate the process of creating as-built BIMs. We sub-divide the overall process into three core operations: geometric modeling, object recognition, and object relationship modeling. We survey the state-of-the-art methods for each operation and discuss their potential application to automated as-built BIM creation. We also outline the main methods used by these algorithms for representing knowledge about shape, identity, and relationships. In addition, we formalize the possible variations of the overall as-built BIM creation problem and outline performance evaluation measures for comparing as-built BIM creation algorithms and tracking progress of the field. Finally, we identify and discuss technology gaps that need to be addressed in future research.

/pdf/automatic-reconstruction-of-as-built-building-information-58eqr9ulay.pdf

Automatic Reconstruction of As-Built Building Information Models from Laser-Scanned Point Clouds: A Review of Related Techniques | NIST

As mobile robotics is gradually moving towards a level of semantic environment understanding, robust 3D object recognition plays an increasingly important role. One of the most crucial prerequisites for object recognition is a set of fast algorithms for geometry segmentation and extraction, which in turn rely on surface normal vectors as a fundamental feature. Although there exists a plethora of different approaches for estimating normal vectors from 3D point clouds, it is largely unclear which methods are preferable for online processing on a mobile robot. This paper presents a detailed analysis and comparison of existing methods for surface normal estimation with a special emphasis on the trade-off between quality and speed. The study sheds light on the computational complexity as well as the qualitative differences between methods and provides guidelines on choosing the ‘right’ algorithm for the robotics practitioner. The robustness of the methods with respect to noise and neighborhood size is analyzed. All algorithms are benchmarked with simulated as well as real 3D laser data obtained from a mobile robot.

Comparison of surface normal estimation methods for range sensing applications

In this work, a framework for motion prediction of vehicles and safety assessment of traffic scenes is presented. The developed framework can be used for driver assistant systems as well as for autonomous driving applications. In order to assess the safety of the future trajectories of the vehicle, these systems require a prediction of the future motion of all traffic participants. As the traffic participants have a mutual influence on each other, the interaction of them is explicitly considered in this framework, which is inspired by an optimization problem. Taking the mutual influence of traffic participants into account, this framework differs from the existing approaches which consider the interaction only insufficiently, suffering reliability in real traffic scenes. For motion prediction, the collision probability of a vehicle performing a certain maneuver, is computed. Based on the safety evaluation and the assumption that drivers avoid collisions, the prediction is realized. Simulation scenarios and real-world results show the functionality.

/pdf/interactive-scene-prediction-for-automotive-applications-3j6aw7zbas.pdf

Interactive scene prediction for automotive applications

This paper presents a probabilistic framework for reasoning about the safety of robot trajectories in dynamic and uncertain environments with imperfect information about the future motion of surrounding objects. For safety assessment, the overall collision probability is used to rank candidate trajectories by considering the probability of colliding with known objects as well as the estimated collision probability beyond the planning horizon. In addition, we introduce a safety assessment cost metric, the probabilistic collision cost, which considers the relative speeds and masses of multiple moving objects in which the robot may possibly collide with. The collision probabilities with other objects are estimated by probabilistic reasoning about their future motion trajectories as well as the ability of the robot to avoid them. The results are integrated into a navigation framework that generates and selects trajectories that strive to maximize safety while minimizing the time to reach a goal location. An example implementation of the proposed framework is applied to simulation scenarios, that explores some of the inherent computational trade-offs.

Safety assessment of robot trajectories for navigation in uncertain and dynamic environments

The verification of evasive maneuvers for autonomous vehicles driving with constant velocity is considered. Modeling uncertainties, uncertain measurements, and disturbances can cause substantial deviations from an initially planned evasive maneuver. From this follows that the maneuver, which is safe under perfect conditions, might become unsafe. In this work, the possible set of deviations is computed with methods from reachability analysis, which allows to verify evasive maneuvers under consideration of the mentioned uncertainties. Since the presented approach has a short response time, it can be applied for real time safety decisions. The methods are presented for a numerical example where two autonomous cars plan a coordinated evasive maneuver in order to prevent a collision with a wrong-way driver.

/pdf/safety-verification-of-autonomous-vehicles-for-coordinated-mi2c19b67i.pdf

Safety verification of autonomous vehicles for coordinated evasive maneuvers

For path planning algorithms of robots it is important that the robot does not reach a state of inevitable collision. In crowded environments with many humans or robots, the set of possible inevitable collision states (ICS) is often unacceptably high, such that the robot has to stop and wait in too many situations. For this reason, the concept of ICS is extended to probabilistic collision states (PCS), which estimates the collision probability for a given state. This allows to efficiently run planning algorithms through crowded environments when accepting a certain collision probability. A further novelty is that the obstacles possibly react to the robot in order to mitigate the risk of a collision. The results show a significant difference in interaction behavior. Thus, this approach is especially suited for active and non-deterministic moving obstacles in the robot workspace.

/pdf/probabilistic-collision-state-checker-for-crowded-a0qy75lm4j.pdf

Daniel Althoff

Papers

Comparison of surface normal estimation methods for range sensing applications

Interactive scene prediction for automotive applications

Safety assessment of robot trajectories for navigation in uncertain and dynamic environments

Safety verification of autonomous vehicles for coordinated evasive maneuvers

Probabilistic collision state checker for crowded environments