By Javier Izquierdo Reyes
Outreach Article

While autonomous driving has been a theme originally raised in science fiction, today it is a reality. Although it is still not possible for a vehicle to drive autonomously without a driver, today it is more common to hear about this technology, thanks to companies such as Tesla or Faraday.

But what is an autonomous vehicle and what types of autonomy exist?

Basically, a vehicle with some degree of autonomy that has the ability to take control of certain functions that would traditionally be performed by a human driver. However, the Society of Automotive Engineers (SAE) have established a frame of reference to locate the level of autonomy that a car can have, which can go from level 0, to level 5 (still non-existent).
Figure 1 shows the six different types of autonomy established and agreed by the SAE to classify autonomous vehicles.

Figure 1. Levels of autonomy according to the Society of Automotive Engineers.

A bit of history regarding autonomous driving

In recent years, there has been a renewed interest in making autonomous vehicles a tangible reality in the world’s vehicle fleet. This is due to the fact that computing power has increased in recent years, which allows for more information processing in less time.

The first approaches to autonomous driving started in the 80s. One of the first efforts was the autonomous land vehicle based on a neural network [1] (ALVINN was its acronym). In 2004, the Department of Defense of the United States promoted the DARPA Grand Challenge, in which the development of autonomous technology was strongly promoted, with Stanford University winning the first event.

Currently, the automotive industry has ventured into the research and development of this technology, however, companies such as Uber or Google through its subsidiary Waymo, are developing both hardware and software for autonomous vehicles, in trend with autonomous technology being a field that has sparked the generation of new companies around the world.

First steps of autonomous driving in Mexico

In Mexico, autonomous driving systems are little known by the general public because level 3 autonomy technology vehicles are too expensive and there is not a great offer yet. However, more and more companies are introducing their vehicles with level 2 autonomy, including greater assistance systems for the driver, such as fatigue detection systems, emergency self-braking, intelligent speed control, among others.

On the other hand, local legislation has not encouraged the implementation of higher levels of technology, or safety, in vehicles sold in the Mexican market, contrary to what occurs in the global market, especially in countries such as the United States or the European Union.

From the technological point of view, autonomous driving is a challenge that can be solved with existing technologies such as cameras, radars, lasers, GPS, and so on. And, although the country’s infrastructure represents another great challenge, Mexican engineering in Artificial Intelligence may be able to achieve a level 4 of autonomous driving.

The opportunity for Mexico

An autonomous vehicle test was carried out by Dr. Raúl Rojas, who, together with his team, developed a vehicle at the Free University of Berlin [2] that was tested on a trip from Nogales, Sonora to Mexico City in a totally autonomous way, reaching speeds of up to 170 kilometers per hour. In an effort to promote and adapt smart vehicles to our country, Dr. Raúl Rojas gave, to several Mexican universities, free prototypes at scale for tests and development of algorithms in order to enable Mexican technology and new companies to grow.

Universities and research centers in Mexico are creating clusters to compete globally in this area. Tecnológico de Monterrey is carrying out various research and development projects in its different campuses. In Mexico City, for example, the ADMAS project [3] seeks to improve and adapt smart vehicles’ systems to the Mexican market, as well as to understand the acceptance of such technology by Mexican users.

While most automotive companies have announced the launch of their autonomous prototypes for 2020 in developed countries, the reality is that there will be a gradual transition in countries like Mexico, where massive use of autonomous vehicles will take more time. However, hardware and software development, or the paradigm shift in terms of the automobile use opens a wide range of possibilities that can be exploited by Mexico, the seventh vehicle assembly power in the world.
The author of this article is an Engineer in Communications and Electronics from the Instituto Politécnico Nacional. He is currently studying his PhD in Engineering Sciences at Tecnológico de Monterrey, Campus Ciudad de México. He is  a professor and coordinator of Borregos CCM Electratón and EcoShell racing teams, and he is in charge of the ADMAS Project, with an interest in robotic vision, machine learning, telecommunications and mechanisms applied to smart vehicles. He has participated in various national and international forums and congresses.

[1] D. a Pomerleau, “Alvinn: An autonomous land vehicle in a neural network,” Adv. Neural Inf. Process. Syst. 1, pp. 305-313, 1989.

[2] R. Rojas and J. Rojo, “Spirit of Berlin: An Autonomous Car for the DARPA Urban Challenge Hardware and Software Architecture,” retrieved Jan, pp. 1-25, 2007.

[3] J. Izquierdo-Reyes, R. A. Ramirez-Mendoza, M. R. Bustamante-Bello, S. Navarro-Tuch, and R. Avila-Vazquez, “Advanced driver monitoring for assistance system (ADMAS),” Int. J. Interact. Des. Manuf., Sep. 2016.


Por favor agrega un comentario!
Favor de ingresar tu nombre