Rainer Stahlmann

The Car as an Ambient Sensing Platform [Point of View]

In recent years, cars have evolved from purely mechanical to veritable cyberphysical systems that generate large amounts of real-time data. These data are instrumental to the proper working of the vehicle itself, but make them amenable to a multitude of other uses. For instance, GPS information has recently been used for a large number of mobility studies in the academic community [1] , [5] , as well as to feed traffic apps such as Google Traffic and Waze. This use of vehicle data is already having a profound impact in science, industry, economy, and society at large. Now, imagine that instead of accessing one single source of vehicle-generated data (GPS), one can access the entire wealth of data exchanged on the controller area network (CAN) bus in near real timeamounting to over 4000 signals sampled at high frequency, corresponding to a few gigabytes of data per hour. What would be the implications, opportunities, and challenges sparked by this transition?

Driver identification using automobile sensor data from a single turn

As automotive electronics continue to advance, cars are becoming more and more reliant on sensors to perform everyday driving operations. These sensors are omnipresent and help the car navigate, reduce accidents, and provide comfortable rides. However, they can also be used to learn about the drivers themselves. In this paper, we propose a method to predict, from sensor data collected at a single turn, the identity of a driver out of a given set of individuals. We cast the problem in terms of time series classification, where our dataset contains sensor readings at one turn, repeated several times by multiple drivers. We build a classifier to find unique patterns in each individuals driving style, which are visible in the data even on such a short road segment. To test our approach, we analyze a new dataset collected by AUDI AG and Audi Electronics Venture, where a fleet of test vehicles was equipped with automotive data loggers storing all sensor readings on real roads. We show that turns are particularly well-suited for detecting variations across drivers, especially when compared to straightaways. We then focus on the 12 most frequently made turns in the dataset, which include rural, urban, highway on-ramps, and more, obtaining accurate identification results and learning useful insights about driver behavior in a variety of settings.

Technical evaluation of GLOSA systems and results from the field

The goal of Green Light Optimal Speed Advisory (GLOSA) systems is to lower CO2 emissions and to avoid unnecessary stopping in intersection approach scenarios by giving speed advices to drivers based on current and future traffic light signal phase timings. These systems have been widely evaluated by means of simulation and, while most research focuses on the impact assessment of GLOSA along with environmental influences, minor attention was drawn to the holistic technical evaluation of included sub-modules and implementations. In this paper we address this problem with a novel and holistic concept for the technical evaluation of IEEE 802.11p based GLOSA systems. We introduce metrics to cover the whole spectrum of GLOSA operations and identify important factors that are usually not considered in simulations, yet, strongly influencing the results. We demonstrate how this concept is used to evaluate the real-world GLOSA system tested in the European Commission co-funded field trial DRIVE C2X. Results derived from Field Operational Test (FOT) data show that our metrics are well-suited to assess the performance of the GLOSA system, but also to identify sources of potential problems or bottlenecks. Based on our findings, we argue that most simulation studies are too optimistic and that further considerations are required to deploy real-world GLOSA systems.

Exploring GLOSA systems in the field: Technical evaluation and results

Green Light Optimal Speed Advisory (GLOSA) systems are believed to be able to lower CO2 emissions, fuel consumption, and travel times by avoiding unnecessary stopping at intersections. Approaching vehicles are given speed recommendations based on current and future traffic light signal phase timings. These systems have been widely evaluated by means of simulation and, while most research focuses on the impact assessment of GLOSA along with environmental influences, minor attention was drawn to the holistic technical evaluation of included sub-modules and implementations.In this extended version of our IEEE VNC 2016 publication, we present a holistic concept for the technical evaluation of IEEE 802.11p-based GLOSA systems. We first give a comprehensive survey on GLOSA systems and studies all around the world and identify remaining problems. We introduce metrics to cover the whole spectrum of GLOSA operations and particularly focus on (modeling) problems we encountered in the field that are often not taken into consideration in simulation studies. We demonstrate how this concept can be used to evaluate the real-world GLOSA system tested in the European Commission co-funded field trial DRIVEC2X. Results derived from Field Operational Test (FOT) data show that our metrics are well-suited to assess the performance of the GLOSA system, and also to identify sources of potential problems or bottlenecks.Based on our findings, we argue that most GLOSA simulation studies are too optimistic in terms of communication performance. Lastly, we give recommendations on how real-world GLOSA systems can be further improved to support a sufficient level of performance.