AI & Machine Learning: The Future Of Transportation
If we talk about the transportation industry, the future is here. In fact, its technology has been around since the 1950s- the years of our first encounter with Artificial Intelligence.
Today, with faster than ever response time benefitting drivers, passengers, and pedestrians altogether, AI and ML are changing transportation and moving towards the future at lightning speed. The idea is not limited to labs anymore but is now being seen on roads. In fact, people on the streets of Guangzhou are even trying to test these vehicles by aggressively cutting in front of them! Not just in China, autonomous vehicles have even already made their way into cities across America including Los Angeles where Google has tested its fleet of self-driving cars for over six years now (and with zero crashes).
Today, the global smart transportation market which, according to MarketsandMarkets’ estimates was valued at $94.5 billion in 2020 is expected to reach $156.5 billion by 2025 at a GAGR of a whopping 10.6 per cent. Artificial intelligence is no more a futuristic philosophy that had little to do with reality. Today, companies are trying to solve real-time problems in the transportation industry using AI.
AI is already being used for and advancing in the following parts of autonomous driving-
Self-diagnosis in vehicles
We already have a convention diagnostic system. However, it is challenging to precisely identify the part at fault. As a result, the vehicle is required to check back at the nearest service center.
LG Electronics is researching an artificial intelligence device that will gather sensor data, perform self-diagnosis on the damaged part, and outputs a self-diagnosis result as a visual warning, an auditory warning, or a haptic warning. Can you imagine how fast it would set the repairing of the vehicle?.
Advanced obstacle detection in autonomous vehicles
Currently, autonomous vehicles use sensors to scan the area for potential obstructions. However, these sensors struggle in scanning a driving environment because of a variety of barriers, unidentified objects, and perceived limitations of vehicle sensor systems. Toyota is determined to bring a solution to this by using an optical model, a neural network model, and a monocular camera to detect barriers linked to unidentified things on the road by calculating the height and distance to an object using a monocular depth estimation. They are also seeking patent protection for the same (US20220057806A1).
Voice assistance for the driver notifying lane change, initiating calls or providing information on vehicle conditions.
Voice assistants have become a household need today. But now we have it in our vehicles too, performing a variety of tasks, such as initiating a call, switching radio stations, or providing information on vehicle conditions. While most vehicles depend on google assistant Fiat Chrysler went a step ahead and partnered with Amazon to create its own digital assistant built on top of Alexa Custom Assistant, leveraging the same technology already used by Amazon Echo smart speakers.
Monitoring the driver
AI face recognition sensors are another thing that’s advancing in the transportation industry. Subaru’s in-car camera recognises the driver’s face as soon as they get into the car. Not only does it adjust the car environment according to the driver but keeps a keen eye on driver behavior thus alerting them if they are distracted or showing signs of drowsiness.
Companies are collaborating with universities to research machine learning and automated driving. One such example is Denso Corp. working with the University of Michigan. The duo has come up with a technology that can predict aggressive driving.
Recognising traffic signals and automating stops based on traffic information
Artificial intelligence not only has the ability to learn and adapt but also makes decisions accordingly. The challenges it majorly faces are due to the limitations in its position accuracy.
Mercedes Benz came up with a novel idea where a system for determining the status of traffic signals may have an onboard camera that is at least capable of taking and identifying pictures of a traffic light in front of a moving vehicle.
In addition, the radar sensor unit and/or lidar sensors are designed to detect information about the conduct of other traffic participants as part of the system for determining the traffic signal light condition and a machine learning algorithm model based on the detection output of the vehicle sensor may be used to obtain the behavior information of the other traffic participants.
Tire performance prediction
Another industry or rather a sub-part of the transportation industry that AI has taken by storm is the tire industry. Companies like Apple, Qualcomm, IBM, etc are working on the digitisation of the tire industry. Working along with start-ups to innovate from embedded sensors to recycling methodologies. One of the startups, Revvo has developed a sensor device for monitoring tire condition, and analyzing tire characteristics, and is coupled with an antenna, which wirelessly communicates with external devices. Revvo’s sensor-enabled artificial intelligence platform gathers this data and provides accurate predictions contributing to increased vehicle uptimes, optimised efficiency, and improved safety.
This is just the start of it. Many car manufacturers are collaborating with autonomous vehicle startups to implement such advanced AI assistance in their vehicles. Companies are working on simulators to train AI-based drivers. Waabi, a Toronto-based AI startup, claims that its advanced simulator can train autonomous vehicles to take about limitless types of driving conditions–in a virtual world–and do so faster and more comprehensively than conventional road tests.
AI and ML are bringing a revolution in the transportation industry. It’ll be fascinating to see how the industry unlocks the full potential of these technologies and fast-forward into the future of automobiles.
