A Fleet Learning Architecture for Enhanced Behavior Predictions during Challenging External Conditions

Abstract

Already today, driver assistance systems help to make daily traffic more comfortable and safer. However, there are still situations that are quite rare but are hard to handle at the same time. In order to cope with these situations and to bridge the gap towards fully automated driving, it becomes necessary to not only collect enormous amounts of data but rather the right ones. This data can be used to develop and validate the systems through machine learning and simulation pipelines. Along this line this paper presents a fleet learningbased architecture that enables continuous improvements of systems predicting the movement of surrounding traffic participants. Moreover, the presented architecture is applied to a testing vehicle in order to prove the fundamental feasibility of the system. Finally, it is shown that the system collects meaningful data which are helpful to improve the underlying prediction systems.

Predicting the Time Until a Vehicle Changes the Lane Using LSTM-based Recurrent Neural Networks

“…In opposition to the so far mentioned machine-learning based approaches, [1] introduced the notion 'physics-based' approaches. Such approaches mostly depend on the laws of physics and can be described with simple models such as constant velocity or constant acceleration [7]. Two wellknown and more advanced model-based approaches are the 'Intelligent Driver Model' (IDM) [8] and 'Minimizing Over-all Braking Induced by Lane Changes' (MOBIL) approach [9].…”

Fleet learning of thermal error compensation in machine tools

“…A Fleet Learning Architecture (FLA) for driver assistance systems under challenging external conditions is presented by Wirthmüller et al [4]. The data in such an architecture is collected from the fleet of vehicles and not only from the initial testing vehicle at the beginning of the process.…”