Benodigde kennis

De benodigde kennis wordt in samenwerking bepaald tijdens werkbijeenkomsten en Ronde Tafels. Dit leidt tot kennisvragen die worden beheerd door het Ministerie van Infrastructuur en Milieu (DGB).

De bovenstaande vragen zijn al deels beantwoord. De antwoorden staan in de documenten uit de collectie en zijn hier verzameld. (stand per maart 2017) Voor meer vragen en antwoorden: zie het Kennisjaarverslag.

TECHNICAL-Architectuur

100% Ja dat kan; nnLearning to Drive: Perception for Autonomous Cars Chapter 3 smooth road detectionnn‘’Accurate perception is a principal challenge of autonomous off-road driving. Perceptive technologies generally focus on obstacle avoidance. However, at high speed, terrain roughness is also important to control shock the vehicle experiences. The accuracy required to detect rough terrain is significantly greater than that necessary for obstacle avoidance. self-supervised learning approach for estimating the roughness of outdoor terrain. Our main application is the detection of small discontinuities that are likely to create significant shock for a high-speed robotic vehicle. By slowing, the vehicle can reduce the shock it experiences. Estimating roughness demands the detection of very small surface discontinuities – often a few centimeters. Thus the problem is significantly more challenging than finding obstacles.’’ Experimental results from this paper show that speed control – reduction  used offers significant improvement in shock  detection.nn80% Self-supervised Road Detection in Desert TerrainnnThe paper presents a method for identifying drivable surfaces in difficult unpaved and offroad terrain conditions as encountered in the DARPA Grand Challenge robot race. Instead of relying on a static, pre-computed road appearance model, this method adjusts its model to changing environments. It achieves robustness by combining sensor information from a laser range finder, a pose estimation system and a color camera. Using the first two modalities, the  system first identifies a nearby patch of drivable surface. Computer Vision then takes this patch and uses it to construct appearance models to find drivable surface outward into the far range. This information is put into a drivability map for the vehicle path planner.

50% Do Autonomous Vehicles Learn?nnSelf learning capabilities of autonomous vehicles. The paper discusses discuss why, whether, and with which challenges and approaches machine learning is possible in its current form in autonomous driving. The paper portrays the view of vehicle technology in particular on this question, and is based on experience from the literature for the area of machine learning.

60%nnModel Based Vehicle Detection and Tracking for Autonomous Urban DrivingnnThis paper describes the moving vehicle detection and tracking module that we developed for our autonomous driving robot Junior. The module provides reliable detection and tracking of moving vehicles from a high-speed moving platform using laser range finders. The paper presents the notion of motion evidence, which allows us to overcome the low signal-to-noise ratio that arises during rapid detection of moving vehicles in noisy urban environments.nn20% Toyota and a Boy Wonder Team Up on Self-Driving CarsnnArticle describing the team up of Toyota and Luminar. Toyota is buying lidar systems from Luminar for their autonomous driving cars. The article says a few things about what lidar systems can see when used in autonoumous driving.nn20% Laser kijkt om de hoeknnArticle describing the tests researchers from Stanford University are doing with lasers. The tests focus on detecting objects behind obstacles, they use a laser to look around the corner. The setup works in real word test with traffic signs and other reflective objects, it isn’t so effective with object that don’t reflect a lot of light. They are improving the system to be able to handle these objects as well as sunny days when there is a lot of interference.


Hieronder een tabel met een aantal documenten die enkel betrekking hebben op het subdomein Technical-Architectuur. Er zijn nog meer documenten gerelateerd aan dit subdomein maar deze zijn ook gelinkt aan andere subdomeinen. De volledige collectie beschikbare documenten in onze online bibliotheek (catalogus en dropbox) 

Titel/Document Author(s)/Schrijvers By Date
Key Considerations in the Development of Driving Automation Systems Andy Christensen, Nissan – North America
Andrew Cunningham, Volkswagen (VW) Group of America Jerry Engelman, Ford Motor Company
Charles Green, General Motors
Charles Kawashima, Mercedes-Benz
Steve Kiger, CAMP
Danil Prokhorov, Toyota Motor Engineering & Manufacturing North America, Inc. Levasseur Tellis, Ford Motor Company
Barbara Wendling, Volkswagen (VW) Group of America
Frank Barickman, National Highway Traffic Safety Administration 06-10-15
An Introduction to Autonomous Control Systems Panos Antsaklis, Kevin Passino, J. Wang Supported in part by the Jet Propulsion Laboratory’ ??/06/1991
Self-supervised Road Detection in Desert Terrain Hendrik Dahlkamp, Sebastian Thrun Stanford University, Intel Corporation ??/??/2006
Real-time motion planning for agile autonomous vehicles Frazzoli, E. University of Illinois at Urbana / MIT ??/02/2015
Learning to Drive: Perception for Autonomous Cars David Stavens Stanford University ??/05/2010
Model Based Vehicle Detection and Tracking for Autonomous Urban Driving Anna Petrovskaya, Sebastian Thrun Stanford University 01-04-2009
Do Autonomous Vehicles Learn? Walther Wachenfeld, Hermann Winner Springer Berlin Heidelberg 22/05/2016