Event: FISITA World Congress 14-16 September 2021 online
Name: Ángel Losada Arias, Automotive researcher and Ph.D. candidate
Company: Polytechnic University of Madrid
Presentation Title: F2021-PIF-073 Improvement of the AEB activation algorithm based on the pedestrian reaction
Please introduce yourself briefly…
I am an Industrial Technologies Engineer, specialized in Mechanical Engineering at the Polytechnic University of Madrid. I completed the qualifying Master's degree in Industrial Engineering, comprising the second year of the Laurea Magistrale at Politecnico di Milano through an exchange program.
I am currently studying for a Ph.D. in Mechanical Engineering at the Polytechnic University of Madrid while working as an assistant researcher at the Institute of Automobile Research in ADAS systems, intelligent driving, Virtual Reality, and dynamic simulation.
Previously, I collaborated in the Formula 1 team of UPM Racing, being in charge of the design and kinematic and dynamic analysis of the rear suspension of the electric F1 car.
I chose this career because of my great passion for the automotive world and especially for the application of new safety and driving assistance technologies and the growing application of virtual reality in this field, especially after my period at Politecnico di Milano, where I was able to acquire valuable knowledge of virtual reality environment modeling in the Virtual Prototyping department. I am also very interested in design and dynamic simulation, especially after my collaboration with the F1 team at the university.
What you will talk about at this year’s FISITA World Congress?
During my presentation at the congress, I will talk about the optimization of Autonomous Emergency Braking systems (AEB) in commercial vehicles, applying virtual reality techniques to obtain the main behavioral patterns of vulnerable road users (VRUs) when crossing the road, and integrating them into the AEB logic algorithm to define the most feasible areas of improvement.
What benefits will participants gain from your presentation?
With this project, it is possible to detect the main areas for improvement in the modeling of AEB systems, understanding the variables that define the behavior of pedestrians when they encroach the crosswalk as crucial information to be managed by the fusion sensor of the ADAS system during its operation. The use of virtual reality techniques to obtain inferences about how users with similar sociodemographic characteristics may react in the pre-collision phase is crucial not only to define their perception of risk but also to define a wide range of reaction models that must be contemplated by the AEB system itself, which must assess the intentionality of the subject according to the variables that define their behavior.
The use of virtual reality techniques to obtain inferences about how users with similar sociodemographic characteristics may react in the pre-collision phase is crucial
Which trends do you consider as the most important for your domain of research and your geographical region?
The protection of vulnerable road users has been one of the main issues to be addressed in the design of new driver assistance systems, as pedestrian and cyclist identification systems need to be sufficiently proactive to warn the driver of the presence of road users.
Although the DGT (Dirección General de Tráfico, Spain) is currently carrying out epidemiological studies to identify areas of vulnerable users accumulation, hard braking, or where there is a risk of collisions, with the aim of identifying black spots and taking action to enhance road safety, ADAS systems are an essential complement to avoid head-on collisions or reduce the severity of the accident in the event of one occurring.
What challenges do these trends present?
With the detailed study of the pedestrian reaction models of the project we are developing (OPREVU) at INSIA-UPM together with Ce-Dint-UPM, which is the subject of this article, it is plausible to determine which points pedestrians look at during the crossing and the average time they spend looking at them, what their speed is in different phases of the crossing and also when a vehicle suddenly appears, or what their general perception of risk is in situations of this type. These patterns are essential for future awareness campaigns and road safety training for vulnerable road users.
An AEB system that can not only identify a road user but also recognize or interpret their intentionality or movements, will have a greater capacity for anticipation, as well as a more gradual response that favors conservation and less wear and tear on the braking system.
And what opportunities do these trends offer?
According to European Road Safety Observatory, pedestrians comprise around 20% of all road deaths in the European Union, a proportion that has remained stable over the last decade. Moreover, cars account for over 70% of vehicles hitting pedestrians, which means that further development of new VRU protection and driver assistance systems is still needed.
This study focuses on the main areas of improvement of AEB systems, and therefore, continues to integrate the use of virtual reality techniques to fully analyze how the pedestrian interacts with the environment and with other road users when crossing. The next stages of this project also seek to go even further by using Eye Tracking technology to identify the main distraction factors of the pedestrian, as well as the integration of a cyclist model in the virtual scenarios, being the next group of VRUs to be analyzed. For this reason, I reckon this topic to be of interest for the future of mobility in the EU, and in the continued reduction of pedestrian and cyclists accidents through the combined action of new technologies in transport and the focus on how the VRU disengages when crossing under different conditions.
The next stages of this project also seek to go even further by using Eye Tracking technology to identify the main distraction factors of the pedestrian, as well as the integration of a cyclist model in the virtual scenarios
Register for the FISITA World Congress now. There is a limited offer of 500 free sponsored registrations for non-presenting delegates!
About the FISITA World Congress
Since 1947, the FISITA World Congress has been the leading international meeting place for industry experts, engineers and executives to exchange ideas and discuss the trends that drive the automotive industry forward. The theme for the 38th FISITA World Congress is 'From automobile to mobility. New roles. New challenges' and it will take place 14-16 September 2021, with the opportunity for the global audience to join virtually available for the first time. Registration is now open, with the Early Bird and special registration rates available until 31 August 2021.
Content from the 2020 FISITA Web Congress is available within the FISITA Technical Library, while material from previous Congress events back to 1998 is available in the FISITA Store. By the time of this year’s Congress, all the content will be available within the FISITA Technical Library, with free access provided to FISITA members.
Partner with us
The FISITA World Congress is one of the most important and prestigious gatherings of mobility engineers and executives in the automotive and mobility calendar. We can offer a number of exceptional promotion opportunities for organisations, associations and companies enabling them to meet their targeted audience and interact with the key decision-makers in and around FISITA 2021.
There are many opportunities to partner with us at the upcoming Congress. Take a look at the virtual prospectus for the full list and easily book a partnership package or virtual booth, or contact Ms. Marie Březinová, Industry Liaison and Sales Manager, Guarant International spol. s r. o., on brezinova@guarant.cz.
Comentários