AI and Autonomous Trucks — Possible Implications of this Tech

Photo by HiveBoxx on Unsplash

Most individuals think of Self-driving cars when they are asked about AI in transportation. However, the introductions of new Autonomous Trucks have the same impact, if not higher, compared to the impact of Autonomous Cars. In this post, let’s take a deeper look at this innovative technology, as well as the numerous pros and cons associated with this relatively new tech.

The Autonomous Truck market is not controlled by a single company. Instead, it is partially controlled by numerous companies that bring new innovations to the industry in order to stay competitive. One of the most successful and well-known company involved in this field is tusimple, a Chinese self-driving startup backed by UPS and Nvidia. They have already completed numerous routes and working on other new routs in Southwestern United States, especially in Arizona and Texas. There are also numerous other companies involved, including Embark and Kodiak (Refer to the diagram below). They have also planned numerous Autonomous truck implementations and have prototypes of Autonomous trucks which use new hardware and software technologies.

While these companies are not as well-known as the Autonomous car giants such as Waymo and Tesla, they are also growing at a fast rate and are bringing new innovations that will help Autonomous trucks become more widespread.

5 Highest Funded Autonomous Truck Companies (beamberlin.com)

Similar to many Autonomous cars, Autonomous trucks use LIDAR as ‘vision’ for the car, helping it collect data about its surroundings. Waymo, one of the main competitors in the field, use two lidar sensors in order to avoid blind spots in the vision model. The data obtained from these sensors are then fed into a Computer Vision models in order for the computer to process the data and detect nearby objects. One of the previously mentioned Autonomous Truck companies, tusimple, relies on the use of camera sensors instead of LIDAR for navigation and object detection. The data, received by either LIDAR or camera sensors, are then fed into one of numerous neural network and machine learning models in order to determine the best course of action for the vehicle. There are many models that are used by companies, such as YOLO (You Only Look Once) and SIFT (Scale-Invariant Feature Transform) to detect objects, process images, determine future actions, and many more. These models are also constantly improved by making models more accurate, reducing the time needed for the program to run, or both.

According to the NHTSA, there are six levels of Automation that range from level 0 (no autonomy) to level 5 (full automation). Some automation features, such as cruise control, lane keeping assistance, and advanced driver assistance system (ADAS), have already been implemented in most trucks, making modern trucks already partially automated. Currently, most Autonomous trucks are either at level 1 or level 2 as a result of all these automation features that are currently present. Many sources suggest that true automation (level 5) of trucks may take more than a few years as there are a lot of technological hurdles that are still present to overcome.

The five levels of Automation for vehicles (Source: NHTSA)

There are numerous positives that are associated with the implementation of this technology. Those are discussed below:

One of the most commonly cited benefits of Autonomous trucks is the increased safety due to the elimination of human error. According to the NHTSA, there were over 5,000 fatalities in motor accidents that involved at least one truck. Disturbingly, many of these accidents are caused by human error, such as distracted drivers, drunk drivers, texting while driving etc. In fact, estimates by the World Economic Forum suggests that about 90% of road accidents are caused by human error. This means that most of the current causes of accidents can be nearly eliminated through use of Autonomous vehicles, especially Autonomous trucks. Widespread use of Autonomous trucks would greatly reduce this number in a similar manner to autonomous cars.

Another positive inherent in the widespread implementation of Autonomous trucks is higher efficiency and cost savings for the companies using these vehicles. Autonomous trucks are much more fuel efficient than trucks controlled by humans. This is partially due to the fact that every action of the truck is programmed, meaning that it can be made to drive in a manner that saves a lot of time and fuel. In addition, Autonomous trucks can drive for much longer than human drivers. According to the World Economic Forum, Self-driving trucks can drive 78% of the day. This is in sharp contrast to human drivers, who can drive for only 29% of the day. Autonomous trucks reduce or eliminate many of the current factors that limit driving time for current trucks.

Demonstration of Object Detection in Autonomous Vehicles (Source: Wikimedia Commons)

As with every new and impactful technology, there are some drawbacks that are associated with this technology. Since Autonomous trucks employ the same technologies in a similar industry as Autonomous cars, this technology also has similar concerns. One of the most well known concerns associated with Autonomous vehicles, especially Autonomous trucks, is job losses. Truck drivers constitute a large portion of US jobs, and they are the largest jobs in some American states. Many Truck drivers and some experts worry that increasing automation of self driving trucks will lead to heavy job losses for truck drivers across the country, which could lead to immense implications across the country. Although there have been studies arguing that increased efficiency will cause a net gain of trucking jobs, automation still remains as one of the major concerns behind this technology.

Another drawback associated with this technology involve ethical concerns about who will be liable in case of accidents. While self driving trucks will make the roads much safer by reducing motor accidents and fatalities, the crashes that still occur pose an ethical problem as to who will be responsible for the crash. Will it be the programmer? Will it be the company which built the technology?, or will it be someone else?. As of right now, there is no clear answer as to who is responsible and it will more than likely pose and ethical concern as autonomous trucks become more commonplace.

Autonomous trucks have a great potential to affect the future, in both positive and negative ways. While true automation for Autonomous trucks have not been reached yet, current levels of innovation and technological advancement suggest that the technology will be available in a few years. Let us the hope that Autonomous trucks become a reality in the near future by minimizing the drawbacks and maximizing the benefits.

Karthik Bagavathy is a Student Ambassador in the Inspirit AI Student Ambassadors Program. Inspirit AI is a pre-collegiate enrichment program that exposes curious high school students globally to AI through live online classes. Learn more at https://www.inspiritai.com/.

A blog that I created to expand on my passions for Artificial Intelligence and other interesting and futuristic technologies