Documente Academic
Documente Profesional
Documente Cultură
Another study on improving fuel economy of commercial vehicles The techniques used to compare and contrast data on this field are
indicated that driver behavior can be responsible for as much as 35% literature review in order to complete a theoretical survey and a
of consumption variation [5], what reflects directly on the TCO. market research to acquire knowledge of the current established
products.
In this context is verified the need of improved driver assistant
systems for getting higher safety and fuel efficiency, thus, lower
TCO.
Page 1 of 7
Behaviorism An American psychologist, by the 1930s, suggested that feelings and
thoughts should also be considered to the Behaviorism [8]. The
nature of the consequences and their effects on behavior was
John B. Watson nominated as Operant Conditioning, which is a learning mechanism
for new behaviors, where the strength of a behavior is modified by
An American psychologist, who originated Behaviorism after reinforcement or punishment, as the figure 2 describes:
conducting researches on animal behavior, child rearing, and
advertising at Columbia University in 1913. It assumes that behaviors Operant Conditioning features
are responses to the environment stimulus [6].
Instant feedback
Positive Punishment
Negative Punishment
Figure 1. Steps of Pavlov experiment with dogs showing that after Applicable when incorrect behavior was performed by removing desirable
conditioning the dog with both food and bell ring stimulus, the dog presented stimulus that increases such behavior. Also used to decrease occurrence of
the salivation (conditioned response) only with the neutral stimulus. an incorrect response.
(https://www.dogalize.com/2018/04/pavlov-dog-classical-conditioning/, 2018)
Teaching machine
Page 2 of 7
Trending Human Machine Interface (HMI)
Analyzing backward the automotive HMI development, the main
purpose of the first automotive HMI was mechanical related. The
functionalities aimed to providing information about the car, like
speed, gas level and rev. counter. Progressively integrating the radio
and the cellphones. Information about the car and entertainment
functions mixed on the HMI are known as infotainment system [10].
Figure 3. Skinner teaching machine. Inside the box on the left side there are a
disc holder for the questions and answers to be provided to the student; on the
right side there is a paper roll in which the students must write the answer and
if necessary correct it after the feedback.(http://pleonasms.rssing.com/chan-
1613600/all_p240.html, 2018).
After analyzing all this content from Behaviorism, it`s time to assume
that the referential characteristics for a good driver feedback system
should be based on the last topics regarding Classical Conditioning
and Operant Conditioning:
It`s possible to change driver behavior by adding external Figure 4. 1961 Buick Invicta instrument panel. (https://www.hemmings.com/,
stimulus using a proper device especially designed based 20108)
on the following topics. The whole device design should be
take into account once it must keep the driver engaged It’s possible to verify that new interactive in-vehicle functionalities
during the trip and after a while using the system; have been influenced by technologies that customers are already used
to or current trends. Furthermore, it’s normal that the new
Correct behavior might be promoted by adding stimulus functionalities are motivated by the construction of safer, more
based on positive reinforcement and/or removing stimulus efficient, and more comfortable vehicle [11].
based on negative reinforcement;
Current trends are pointing the future of HMI to connected mobiles
devices and cloud connection for cloud-based applications, which
Incorrect behavior might be discouraged by adding will reduce the growing hardware demands. HMI will be dynamically
stimulus based on positive punishment and/or removing adaptive, depending on the drivers’ situation and habits. Something
stimulus based on negative punishment. looking like a smartphone with touch screen display, containing
information of the car, roads and the driver’s personal phone
Real time feedback plays an important role mainly by applications.
placing stimulus to the driver on the go, what enables
stronger association of correct and incorrect driver behavior
compared to after trip feedback systems. Also makes
possible to design real time gaming apps in order to keep
drivers engaged.
Page 3 of 7
As the number of smartphones users increase from 62,6 million in
2010 to 236,8 million in 2019 in the United States [12], Americans
say that “they couldn’t live without their smartphone” [13], mainly
the so-called digital natives [14].
Engaging drivers
Figure 6. Three levels of driving tasks proposed by Rasmussen, Suetomi and
Whereas the number of drivers bring their own devices inside the Stanton.
cars, calling or texting, and accessing social media while driving is
increasing [15], it’s also a good opportunity for safe and fuel save
driving games in order to get their engagement.
Lack of engagement while driving can be caused by unsafe Data input for the driver feedback device
smartphone usage [16]. It was suggested that smartphone apps should
be used to lock the smartphone while the cars are in motion in order The content of the feedback, real time or after trip, that could be
to mitigate risks of distraction [17]. However, adding engaging manipulated to apply reinforcement or punishment Skinner’s concept
stimuli for drivers through gaming apps have a direct impact on road must be based on trustable information. Data input can be streamed
safety and user experience, showing positive effect on driver from basically 3 sources [20]:
behavior [18].
1.Vehicle (acceleration, speed, location, rev counter, etc.)
There are some related applications that provide feedback to drivers
without consider when it’s safe [19]. As result, they take the focus of 2.Road (traffic, topographic, road inclination, etc.)
the driver away and it should be avoided.
3.Driver (heart rate, blood pressure, sleepiness, etc.)
Projecting a system that provide engaging experiences to the drivers
requires equilibrium between safety and fun [20]. Another
Input data from vehicle
perspective of an extra device installed in the vehicle is that it would
require additional attention from the driver, who must keep watching
the road and all the devices. Resulting in increase of workload and Vehicle data can be obtained from a device not connected to the
impacting on the acceptance of the new system [26]. vehicle, just attached to the structure, like a cellphone, which has
been used to get vehicle position, speed, acceleration, deceleration
and deflection angle using only the cellphone sensors: accelerometer,
gyroscope and magnetometer. Using no external sensors in order to
get a more cost efficient, user-friendly and minimalist system as
Safety engaging experience described in [27]. Also in [27], there is a system block diagram which
revels the logical process of the data, as show in figure 7, wich shows
Once driving is a visual task [21], a real time feedback system should the data input block through a preprocessing via smoothing filter
use another way of transmit information. Therefore, sensory and/or block. The following block detect the potential events based on the
auditory ways would be safer than a pure visual one. temporal range of the signals. Next the system algorithms detect the
events like right and left turn based on templates for safe driving
The main tasks of a driver were first defined by Rasmussen in [22] habits, shown in figure 8. After that, once the event matches a
and enhanced by Suetomi [23] and Stanton [24] in three levels. The template, a decision making algorithm classifies between risky and
first tasks level is safety related, thus, the most relevant. It’s about safe driving habits and display. This solution
vehicle guidance and stabilization. The second level is about
enhancing the driving ability, like the ability to shift gear when Other options to input signal data are:
driving a manual transmission vehicle. The third level is the only one
not related to driving. It`s related to distractions while driving, e.g., OEM device connected to the vehicle CAN bus or
using the cellphone and changing radio station. connected to the vehicle sensors;
Devices from tech companies connected to the On-Board
Looking forward to keep driver`s focus on safety driving activities, Diagnostics (OBD) connector via Bluetooth dongles, or
it’s necessary to postpone the non-guidance features of a real time connected to the SAE FMS J1939 CAN bus, or also linked
feedback system while dangerous situations are occurring [25], such to its own sensors;
as harsh braking or fast lane changing. In systems designed to give
feedback after the trip, this wouldn’t be a concern.
Page 4 of 7
Input data from road
In terms of precision and cost benefit, road and vehicle data sources
are the best options at the current technological state. It’s a matter of
work on this sources to acquire as much information as possible. As
will be discussed next, driver source of data is still not accurate at
low prices.
Figure 8. Templates for right and left turn, which is compared to the signal
input in order to identify the vehicle events ([27]).
In order to safely engage drivers to a driver feedback platform, it’s 11. Meixner G. et al. (2017) Retrospective and Future Automotive
recommended to take care about when real time feedback will be Infotainment Systems—100 Years of User Interface Evolution.
displayed, otherwise it will take out attention from the driver. In: Meixner G., Müller C. (eds) Automotive User Interfaces.
Furthermore, it’s important other means of feedback than visual, once Human–Computer Interaction Series. Springer, Cham.
driving is a visual task.
12. eMarketer. (2016). Number of smartphone users in the United
Analyzing the literature, most of related work do not take into States from 2010 to 2019 (in
account how effective is the feedback. They care about the source of millions). http://www.statista.com/statistics/201182/forecast-of-
data, how the data will be processed, how and when the feedback will smartphone-users-in-the-us/.
be displayed, also the quality and cost of the data acquainted.
However, they do not take into account any Behaviorism concept 13. Anderson, M. (2015a). 6 facts about Americans and their
purposely. What reveals that there is space to further works on this smartphones.http://www.pewresearch.org/fact-
field. tank/2015/04/01/6-facts-about-americans-and-their-
smartphones/.
Overall, the analyzed literature showed positive result by the changes
14. Selwyn, N. (2009). The digital native: myth and reality.
in decision making outcomes about driving styles through promoting
In: Publication Cover (pp. 364–379). London: s.n..
awareness.
15. Vollrath, M., Huemer, A. K., Teller, C., Likhacheva, A., &
Fricke, J. (2016). Do German drivers use their smartphones
safely? —Not really! Accident Analysis and Prevention, 96, 29–
References 38.
1. SAE (2014), Levels of Driving Automation Are Defined In New 16. Steinberger, F., Moeller, A., & Schroeter, R. (2016). The
SAE International Standard J3016, Society of Automotive antecedents, experience, and coping strategies of driver boredom
Engineers (www.sae.org); at in young adult males. Journal of Safety
www.sae.org/misc/pdfs/automated_driving.pdf. Research, 59(December), 69–82. ISSN 0022-
4375. http://dx.doi.org/10.1016/j.jsr.2016.10.007.
Page 6 of 7
17. Coben, J. H., & Zhu, M. (2013). Keeping an eye on distracted 31. You, C.-W., Montes-de-Oca, M., Bao, T. J., Lane, N. D., Lu, H.,
driving. JAMA, 309, 877–878. Cardone, G., Torresani, L., & Campbell, A. T. (2012). CarSafe:
A driver safety app that detects dangerous driving behavior.
18. Heslop, S. (2014). Driver boredom: Its individual difference using dual-cameras on smartphones. In Proceedings of the 2012
predictors and behavioural effects. Transportation Research ACM Conference on Ubiquitous Computing (pp. 671–672).
Part F: Traffic Psychology and Behaviour, 22, 159–169. New York, NY, USA: ACM.
19. McCall, R., & Koenig, V. (2012). Gaming concepts and 32. Hong, J.-H., Margines, B., & Dey, A. K. (2014). A smartphone-
incentives to change driver behaviour. In Ad Hoc Networking based sensing platform to model aggressive driving behaviors.
Workshop (Med-Hoc-Net), 2012 The 11th Annual In Proceedings of the 32nd Annual ACM Conference on Human
Mediterranean (pp. 146–151). Factors in Computing Systems (pp. 4047–4056). New York,
NY, USA: ACM.
20. Steinberger F., Schroeter R., Babiac D. (2017) Engaged
Drivers–Safe Drivers: Gathering Real-Time Data from Mobile 33. Aguilar, S. R., Merino, J. L. M., Sánchez, A. M., & Valdivieso,
and Wearable Devices for Safe-Driving Apps. In: Meixner G., Á. S. (2015). Variation of the heartbeat and activity as an
Müller C. (eds) Automotive User Interfaces. Human–Computer indicator of drowsiness at the wheel using a smartwatch.
Interaction Series. Springer, Cham. International Journal of Artificial Intelligence and Interactive
Multimedia 3.
21. Sivak, M. (1996). The information that drivers use: Is it indeed
90% visual? Perception, 25, 1081–1089. 34. Russell, J. A. (1980). A circumplex model of affect. Journal of
Personality and Social Psychology, 39, 1161–1178.
22. J. Rasmussen. Information processing and human-machine
interaction - an approach to cognitive engineering. North
Holland Series in System Science and Engineering, 1986.
27. Eren, H., Makinist, S., Akin, E., & Yilmaz, A. (2012).
Estimating driving behavior by a smartphone. In Intelligent
Vehicles Symposium (IV), 2012 IEEE (pp. 234–239).
30. Schroeter, R., Soro, A., & Rakotonirainy, A. (2013). Social cars:
sensing, gathering, sharing and conveying social cues to road
users. In B. Guo, D. Riboni & P. Hu (Eds.) Creating personal,
social, and urban awareness through pervasive computing (pp.
176–200). IGI Global.
Page 7 of 7