17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 2014
Using current sensing technology, a wealth of data on driving sessions is potentially available t... more Using current sensing technology, a wealth of data on driving sessions is potentially available through a combination of vehicle sensors and drivers' physiology sensors (heart rate, breathing rate, skin temperature, etc.). Our hypothesis is that it should be possible to exploit the combination of time series produced by such multiple sensors during a driving session, in order to (i) learn models of normal driving behaviour, and (ii) use such models to detect important and potentially dangerous deviations from the norm in real-time, and thus enable the generation of appropriate alerts. Crucially, we believe that such models and interventions should and can be personalised and tailor-made for each individual driver. As an initial step towards this goal, in this paper we present techniques for assessing the impact of cognitive distraction on drivers, based on simple time series analysis. We have tested our method on a rich dataset of driving sessions, carried out in a professional simulator, involving a panel of volunteer drivers. Each session included a different type of cognitive distraction, and resulted in multiple time series from a variety of on-board sensors as well as sensors worn by the driver. Crucially, each driver also recorded an initial session with no distractions. In our model, such initial session provides the baseline times series that make it possible to quantitatively assess driver performance under distraction conditions.
Physica A: Statistical Mechanics and its Applications, 2014
ABSTRACT Within this work the Nagel–Schreckenberg (NS) cellular automata is used to simulate a ba... more ABSTRACT Within this work the Nagel–Schreckenberg (NS) cellular automata is used to simulate a basic cyclic road network. Results from SwitchEV, a real world Electric Vehicle trial which has collected more than two years of detailed electric vehicle data, are used to quantify the results of the NS automata, demonstrating similar power consumption behavior to that observed in the experimental results. In particular the efficiency of the electric vehicles reduces as the vehicle density increases, due in part to the reduced efficiency of EVs at low speeds, but also due to the energy consumption inherent in changing speeds. Further work shows the results from introducing spatially restricted speed restriction. In general it can be seen that induced congestion from spatially transient events propagates back through the road network and alters the energy and efficiency profile of the simulated vehicles, both before and after the speed restriction. Vehicles upstream from the restriction show a reduced energy usage and an increased efficiency, and vehicles downstream show an initial large increase in energy usage as they accelerate away from the speed restriction.
Parking Guidance and Information (PGI) System becomes highly favorable for reducing circulating t... more Parking Guidance and Information (PGI) System becomes highly favorable for reducing circulating traffic and making efficient use of existing parking facilities. This paper is to examine the factors influence drivers' willingness to use PGI. Factor analysis and the Structure Equation Model (SEM) were used to identify the latent attitudinal factors and the sensitivity of the factors was judged by Bayesian network. The heterogeneity of the factors was explored based on driver's gender, age, driving years, education and travel frequency. The results show that drivers' willingness to use PGI is significantly correlated to five attitudinal factors: perception of existing PGIs, difficulty in parking, confidence in the accuracy of the information, easy acquisition of information and information attributes. Male drivers, younger drivers, novice drivers and drivers who travel less frequently have lower level of willingness to use PGI.
Conference Record - IEEE Conference on Intelligent Transportation Systems
Within this paper a Markov chain is used to simulate the charging and driving behavior of a cohor... more Within this paper a Markov chain is used to simulate the charging and driving behavior of a cohort of electric vehicles. The probability transition matrix is constructed by analyzing the real world data from an ongoing Electric Vehicle (EV) trial, Switch EV. The vehicles are split into a combination of drive, park and charge states as well as low, mid and high state of charge. The Markov chain is formed from the probability of transitioning between each state for a specified time period. The Markov chain produced results from both work and home based charging regimes that correlate well with the real world results with modeled results showing the features present in real world data. From comparisons of fresh sample runs to the ideal modeled charging distribution it can be seen that the Markov chain has little systematic difference between the ideal distribution and the newly generated distribution within a short period of time. The Markov chain was used to investigate the effect on ...
17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 2014
Using current sensing technology, a wealth of data on driving sessions is potentially available t... more Using current sensing technology, a wealth of data on driving sessions is potentially available through a combination of vehicle sensors and drivers' physiology sensors (heart rate, breathing rate, skin temperature, etc.). Our hypothesis is that it should be possible to exploit the combination of time series produced by such multiple sensors during a driving session, in order to (i) learn models of normal driving behaviour, and (ii) use such models to detect important and potentially dangerous deviations from the norm in real-time, and thus enable the generation of appropriate alerts. Crucially, we believe that such models and interventions should and can be personalised and tailor-made for each individual driver. As an initial step towards this goal, in this paper we present techniques for assessing the impact of cognitive distraction on drivers, based on simple time series analysis. We have tested our method on a rich dataset of driving sessions, carried out in a professional simulator, involving a panel of volunteer drivers. Each session included a different type of cognitive distraction, and resulted in multiple time series from a variety of on-board sensors as well as sensors worn by the driver. Crucially, each driver also recorded an initial session with no distractions. In our model, such initial session provides the baseline times series that make it possible to quantitatively assess driver performance under distraction conditions.
Physica A: Statistical Mechanics and its Applications, 2014
ABSTRACT Within this work the Nagel–Schreckenberg (NS) cellular automata is used to simulate a ba... more ABSTRACT Within this work the Nagel–Schreckenberg (NS) cellular automata is used to simulate a basic cyclic road network. Results from SwitchEV, a real world Electric Vehicle trial which has collected more than two years of detailed electric vehicle data, are used to quantify the results of the NS automata, demonstrating similar power consumption behavior to that observed in the experimental results. In particular the efficiency of the electric vehicles reduces as the vehicle density increases, due in part to the reduced efficiency of EVs at low speeds, but also due to the energy consumption inherent in changing speeds. Further work shows the results from introducing spatially restricted speed restriction. In general it can be seen that induced congestion from spatially transient events propagates back through the road network and alters the energy and efficiency profile of the simulated vehicles, both before and after the speed restriction. Vehicles upstream from the restriction show a reduced energy usage and an increased efficiency, and vehicles downstream show an initial large increase in energy usage as they accelerate away from the speed restriction.
Parking Guidance and Information (PGI) System becomes highly favorable for reducing circulating t... more Parking Guidance and Information (PGI) System becomes highly favorable for reducing circulating traffic and making efficient use of existing parking facilities. This paper is to examine the factors influence drivers' willingness to use PGI. Factor analysis and the Structure Equation Model (SEM) were used to identify the latent attitudinal factors and the sensitivity of the factors was judged by Bayesian network. The heterogeneity of the factors was explored based on driver's gender, age, driving years, education and travel frequency. The results show that drivers' willingness to use PGI is significantly correlated to five attitudinal factors: perception of existing PGIs, difficulty in parking, confidence in the accuracy of the information, easy acquisition of information and information attributes. Male drivers, younger drivers, novice drivers and drivers who travel less frequently have lower level of willingness to use PGI.
Conference Record - IEEE Conference on Intelligent Transportation Systems
Within this paper a Markov chain is used to simulate the charging and driving behavior of a cohor... more Within this paper a Markov chain is used to simulate the charging and driving behavior of a cohort of electric vehicles. The probability transition matrix is constructed by analyzing the real world data from an ongoing Electric Vehicle (EV) trial, Switch EV. The vehicles are split into a combination of drive, park and charge states as well as low, mid and high state of charge. The Markov chain is formed from the probability of transitioning between each state for a specified time period. The Markov chain produced results from both work and home based charging regimes that correlate well with the real world results with modeled results showing the features present in real world data. From comparisons of fresh sample runs to the ideal modeled charging distribution it can be seen that the Markov chain has little systematic difference between the ideal distribution and the newly generated distribution within a short period of time. The Markov chain was used to investigate the effect on ...
Uploads
Papers by Phil Blythe