Autonomous Driving and What to Expect in the Near Future

For decades, humans have dreamed of a life where no driving is required to get to work, where accessibility to drivers is not an issue, where you can guarantee the delivery of goods to any part of the country, where safety is paramount and guaranteed, and the probability for an accident is close to zero. For the past few years, we´ve seen a significant evolution in transportation technology; one in particular holds a great deal of promise to reshape our cities and daily lives: autonomous vehicles. On the one side, we have robotaxis, vehicles designed to transport people intra-city with no driver required, promising to reshape cities of the future. On the other side, autonomous trucks are designed to transport goods around the country safely, regardless of the hour or time of the year. As we edge closer to widespread adoption of these technologies in some countries, it´s crucial to understand their potential impacts, challenges, and varying approaches taken by companies leading the way.

Robotaxis are self-driving vehicles designed to transport people similar to taxis or ride-hailing services without the need of a driver. To operate correctly, it leverages advances in artificial intelligence systems, sophisticated sensor arrays, and complex algorithms to navigate streets and merge with the surrounding chaotic world.

Autonomous trucks, also known as self-driving trucks, use a similar approach and technology to robotaxis but are designed for freight transport. These vehicles have the potential to revolutionize the logistics industry in the near future.

Both solutions and their widespread adoption could have a huge impact on society and how it operates. The first and most important promise is safety. Eliminating human error, which accounts for the vast majority of traffic incidents, could save thousands of lives in cities and on highways. It will also improve accessibility for disabled and elderly individuals. It can change how cities are planned, with reduced need for parking spaces and more efficient traffic. We could see cities change in the near future as a result.

Another factor with potential impact on our economy is that it will reduce the dependency on drivers, allowing us to find them when needed. Since COVID, we´ve seen an increased difficulty in finding drivers to operate trucks, especially in Mexico and the United States. This has created a shortage in the availability of trucks, increasing freight costs and delivery times. Younger generations do not find the life of a truck driver, or spending 15 hours per day driving a vehicle to make a living, attractive.

As the race to develop and deploy autonomous vehicles intensifies, different approaches have emerged for passenger and freight transport.

For the passenger business, two companies lead the way: Tesla and Waymo (Google). Tesla´s approach relies on a vision-based system; this means using only cameras and a complex neural network. The company uses multiple cameras and ultrasonic sensors to achieve what it describes as human-level perception. It gathers vast amounts of data from its vehicles to train its neural networks and then deploy these updates on its vehicles. Tesla also developed its own processing units, optimized for video analysis and vehicle control. This approach does not rely on highly detailed, pre-mapped environments; it is considered a more scalable solution that can adapt to new areas quickly.

Waymo, on the other hand, has taken a more cautious and methodical approach. Its strategy relies heavily on a combination of sensors, including lidar, a laser that continuously scans the environment with high precision, radar, and high-resolution cameras, to create a comprehensive view of the vehicle's surroundings. Waymo emphasizes the importance of using detailed 3D maps for the areas where its vehicles operate to help its system understand the environment with greater precision.

For freight transport, Waymo is also a leader in the market, competing with Tusimple and Aurora. Freight vehicles follow Waymo´s approach of using an extensive number of sensors, including lidar, radar, and high-definition cameras, but there is heavier investment  in longer-range sensing and faster decision-making.

Despite the rapid progress in autonomous vehicle technology for both passenger and freight transport, significant challenges remain. To achieve human-level perception and decision-making in all possible driving scenarios, neural networks (AI systems) need to understand how humans behave in different environments. This is, in my opinion, the most difficult task to achieve full autonomy while driving, especially in developed countries where people do not follow traffic instructions so easily. Governments still have a long way to go in the development of comprehensive laws and regulations governing testing and deployments of autonomous vehicles.

Companies need to gain widespread public acceptance and trust in the safety and reliability of these vehicles. In recent months, we´ve seen people sabotaging autonomous vehicles as a demonstration of mistrust in them. In addition, companies need to show that large trucks sharing the road with passenger vehicles is very safe and key for their adoption. Programmers must consider complex ethical dilemmas in decision-making algorithms — basically, who lives and dies (passenger, pedestrian). To understand more about this, I recommend you visit the MIT moral machine experiment (www.moralmachine.net).

Cybersecurity is also extremely important; we need to ensure that autonomous vehicles are protected against hacking for safety and public confidence, especially for freight vehicles carrying valuable or potentially dangerous cargo. Cities and roads need to change to optimize the performance of these vehicles, both towns and highways.

It is important to understand that the path toward autonomous vehicles is difficult and each country will approach the introduction of these technologies in a different way. There are several other roadblocks that companies must go through that require significant CAPEX investments, mainly on computation. The race for AI, creating more complex large language models (LLM´s), and  to be the first to create an artificial general intelligence has created fierce competition for computational capacity, with billions invested in server farms and increasing energy demand significantly. There is inherent competition for computing allocating between neural networks for autonomous vehicles and LLM´s. 

As we stand on the cusp of this transportation revolution, it's clear that autonomous vehicles have the potential to fundamentally transform both personal mobility and freight transport. The technology promises to make our roads safer, our cities cleaner, and our transportation systems more efficient and accessible.

However, the path to widespread adoption is not without its challenges. As companies continue to refine their approaches and push the boundaries of what's possible, regulators, urban planners, and society at large must grapple with the implications of this technology.

The coming years will be crucial in determining how quickly and smoothly we transition to a future where autonomous vehicles are a common sight on our roads and highways. As this technology continues to evolve, ongoing public discourse, careful regulation, and rigorous safety testing will be essential to ensure that the revolution in autonomous vehicles delivers on its promise of safer, cleaner, and more efficient transportation for both passengers and freight.

I believe in Mexico in particular the adoption of this technology will take more time. From the sidelines, we will see how countries adopt this technology and make decisions on the go, making sure we can access the most safe and reliable technology at the time.