Euro

The european project Training Activities to Implement the Data Protection Reform (TAtoDPR) has received funding from the European Unionís Rights, Equality and Citizenship (REC) Programme of the European Union under Grant Agreement No. 769191

The contents of this Journal represent the views of the author only and are his/her sole responsibility. The European Commission does not accept any responsibility for use that may be made of the information it contains.

Home / ISSUES / Issue / Empirical methodologies for the design of innovative autonomous driving solutions

back print content read pdf content


Empirical methodologies for the design of innovative autonomous driving solutions

Anna Irene Cesarano

Abstract: This article aims to present itself as an overview of the world of self-driving cars, giving a glimpse albeit fleeting to my research project and clarifying some basic concepts to understand the future developments of this discipline. Concepts like automation levels, self-driving cars, artificial intelligence are essentials and most important to understand all the innovative flow of autonomous driving. Autonomous driving represents the challenge of the future, although many critical aspects must be overcome and addressed.

 

Key-words: Self-driving cars, Artificial intelligence, Automation levels, Driverless cars, Research project, Autonomous cars.

 

Summary:  1. Self-driving cars. - 2. Automation levels: SAE’s classification. - 3. My research project. - Conclusions.

 

   

Summary:

1. Self-driving cars - 2. Automation levels: SAEís classification - 3. My research project - 4. Conclusions - Notes


1. Self-driving cars

The automotive sector is constantly evolving, self-driving cars are the result of a complex design that adopts a good variety of devices and sensors[1] that capture information from the external environment which is then transmitted to an internal computer, in order to guarantee the vehicle efficiency, safety, stability for both passengers and for people who are near the car itself.Self-driving cars[2] can be defined as those vehicles capable of autonomously taking control of the car and coping with all aspects of driving such as accelerating and braking, steering control, gear shifting. Self-driving cars, or in other names like “autonomous cars” “driverless cars” are the product of the continuous evolution of the automotive sector and of a new concept of transport and mobility, which through the application of artificial intelligence become "smart". The modern concept of autonomous driving with cars controlled by artificial intelligence and equipped with the most varied and modern sensors and technological devices, is able to disrupt the classic concept of the car, paving the way for new lines of research and more and more design techniques user centered           (UCD), for safer and smarter mobility.[3] 


2. Automation levels: SAEís classification

SAE[4] society of Automotive Engineers, a body responsible for regulation in the aerospace, automotive and vehicle sector[5], made up of scientists, engineers and scholars, has classified six levels of autonomous driving, which start from level 0, the lowest, to the highest. level 5, although the rapid development of the automotive market suggests a future increase in levels.

Specifically, the classification of the levels is as follows:

Level 0, No automation: this is therefore the traditional car, in which the driver controls the car without any type of support from a driver assistance system.

Level 1, Driving assistance: where, thanks to on-board electronics, the driver is helped by information but still has full control of the car in terms of driving (and therefore in acceleration, braking, steering and so on Street). On a "legal" level, if we can say so, the driver is still fully responsible for everything that happens. The car simply gives support to the driver by capturing information from the outside worlds such as obstacles, dangerous situations and so on. Systems such as ABS, Cruise Control or AirBag, parking sensors are Level 1.

Level 2, Partial automation: in Level 2 of assisted driving the electronics start to integrate into the guide allowing it to intervene in certain situations. There is therefore no full control of the vehicle by the driver, who must however continue to deal with driving in almost all its aspects. This type of automation intervenes only in some cases such as, for example, assisted braking (Brake Assist) following the detection of an object that is on a collision course with our car, or emergency collision braking. Another example may be Lane Assist once again, where instead of receiving only an acoustic signal, the car will correct the steering angle to return to the lane.Level 3[6], Limited autonomous driving: With this assistance level, in fact, the electronics are able to automate driving in the fundamental aspects: acceleration, braking, steering. Obviously, in addition to "basic" driving, the car also has all the safety and automation levels described above, but the driver must still keep his eyes open, ready to intervene in adverse circumstances such as unfavorable environmental conditions, unpaved road surface or particularly disconnected and so on. Many cars have this level of automation and, for example, the Park Assist with which the car parks itself, is part of this level.Level 4, High automation: In Level 4 of assisted driving, on the other hand, you can relax and let the car take you to your destination without intervening because the electronics are able to manage any situation that happens in front of them. However, although the car can manage driving completely, the driver cannot use it in adverse conditions. In these cases, the car will ask the driver to start driving again. Level 5[7], Complete automation: one can speak explicitly of autonomous driving in all its aspects. In this case, therefore, human intervention is not required in any situation and the car can also arrive at its destination without someone in the passenger compartment, making decisions and choosing the best route based on traffic. Autonomous vehicles scan the environment with techniques such as radar, lidar, GPS, and artificial vision. Advanced control systems interpret the information received to identify appropriate routes, obstacles and relevant signs.[8]


3. My research project

My research project[9] involves the development of innovative systems to support autonomous driving (partially autonomous driving), combining one of the transversal sectors (Information & Communication Technologies) with a vertical sector (Transport and Logistics) of particular relevance for the Campania. In particular, the research theme involves the development of a new safer and smarter mobility paradigm, with increasing automation shares that must be harmonized with human characteristics to allow it to maintain an active role in the new socio-technical system. The research theme intends to investigate the methodologies necessary for the conceptualization, prototyping and verification of more suitable user interfaces to ensure effective interaction with the driver (in the context of partially or totally autonomous driving), taking into account technological aspects such as those relating to the individual and his cognitive, behavioral, etc. The application of innovative design methodologies (based on UCD and UX) will guide the design in the conceptualization phase, to allow a preliminary prototyping of the technological construct and its interface before it becomes a definitive product and allow to verify in advance how the people interact with the technological system. If this cooperation[10], studied through empirical methodologies, gives promising signals or highlights some critical elements, the user interface may be modified accordingly. Therefore, in addition to the conceptualization and prototyping methodologies, the development of technology verification methods are also an essential part of the innovative research system.Simulation environments are available within the Scienza Nuova[11] research center, both aimed at the automotive domain, and reconfigurable thanks to the presence of a driving simulator and a virtual environment in which different interaction experiences can be reproduced. Several technologically advanced tools can be used such as eye-tracking, biometric sensors, systems for detecting the emotional state through videocameras etc.


4. Conclusions

The automotive sector shows an innovative and technological character in its continuous evolution. The classic concepts of driving, car, driver are supplanted in favor of a new concept of safer and smarter mobility. Autonomous driving represents the challenge of the future, although many critical aspects[12] must be overcome and addressed. Autonomous driving or robotaxis could be able to revolutionize many sectors of community life, from the purely social one such as allowing certain types of people to increase their productivity or even disabled or elderly people to move in comfort. While on the purely economic side, autonomous driving would be able to save on the costs of resources and staff employed, but the real turning point was that it would be able to drastically reduce road accidents, which according to some statistical data would be caused in the 95%[13] of cases from human distractions, saving many lives.  


Notes

[1] Sensors and devices are: lidar, radar, videocameras, artificial vision, ultrasound.

[2] P. Koopman, M. Wagner, ‘Autonomous Vehicle: An Interdisciplinary Challenge’ [2017] IEEE Intelligent Transportation Systems Magazine, 90-96.

 

[3] P. Koopman, M. Wagner, ‘Challenges in Autonomous Vehicle Testing and Validation’ (2016) 4(1) SAE International   Journal of Transportation. Safety, 15-24, , https://doi.org/10.4271/2016-01-0128., SAE World Congress Exhibition

[4] SAE J3018, Mar. 20015 Guidelines for Safe On-Road Testing of SAE Level 3, 4, and 5 Prototype Automated Driving

Systems (ADS); SAE J3061, Jan 2016 Surface Vehicle Recommended Practice: Cybersecurity Guidebook for Cyber-Physical Vehicle Systems.

Cfr. A. Cioffi, Digital Strategy. Strategie per un efficace posizionamento sui canali digitali (Hoepli, 2018). https://smartrider.ch/it/attualita/5-livelli-di-automazione-sae

[5] A. Cioffi, Digital Strategy ibid.

 

[6] Currently level 3 is what is the highest level of automation on the market

[7] A. Semoli, AI marketing. Capire l’intelligenza artificiale per cogliere le opportunità (Hoepli, 2019).

[8] For more in-depth legal analysis see M. C. Gaeta, Liability rules and self-driving cars: The evolution of tort law in the light of new technologies (Editoriale scientifica, 2019).

[9] My tutor is Roberto Montanari, professor PHD HMI and manager RE:Lab the Interaction Engineering Company

[10] The concept of cooperation between car and driver is defined in an innovative way by TeamMate car in TeamMateHMI design, implementation and V&V results from 1st cycle, 30-06-2017, authors R. Montanari et al., project number: 690705, www.automate-project.eu

[11] Scienza nuova is the research center of University Suor Orsola Benincasa

[12] Many aspects must be considered as people's distrust and fear of autonomous driving

[13] https://www.europarl.europa.eu/news/it/headlines/society/20190410STO36615/le-statistiche-sugli-incidenti-stradali-mortali-nell-ue-infografica


  • Giappichelli Social