AI in Transportation: Market Overview and Applications

Introduction

With the concerns surrounding artificial intelligence (AI) and its capacity to do harm – from eliminating jobs to eliminating civilization – the application of AI to transportation, specifically, ground transportation, offers considerable promise for a safer future. In the field of autonomous, self-driving vehicles, AI will still eliminate jobs – cabbies and truckers, for example – but could render our roads and highways safer by dramatically reducing the incidence and severity of traffic accidents, improving not only vehicle management but traffic and highway management as well.

According to the Traffic Safety Facts Annual Report, in the US in 2021, there were:

• 6,102,936 police-reported motor vehicle traffic crashes, resulting in
• 42,939 fatalities, and
• 2,497,657 injuries.

Experiencing far fewer vehicular fatalities and injuries is the reason why AI in transportation is being embraced. The transportation sector is endeavoring to make travel safer, healthier (with the transition to electric vehicles and the elimination of greenhouse gas emissions), and more enjoyable for drivers, passengers, and pedestrians.

The AI in Transportation Market

According to the Business Research Company, a global market research and consulting firm, the global AI in transportation market, valued at about $3.25 billion in 2023, is expected to reach $6.3 billion in 2027, realizing a compound annual growth rate (CAGR) of 18.2 percent.

Prominent players in the AI in Transportation space include:

• Volvo
• Mercedes-Benz
• Alphabet (Google)
• Intel
• NVIDIA
• Valeo
• ZF
• Continental
• Magna
• Microsoft
• PACCAR
• Bosch
• Scania
• Xevo
• Zonar²

Autonomous Vehicles

Having long been the dream of every tired and frustrated driver, autonomous or self-driving vehicles, when fully realized, can provide massive benefits. Not only can they offer additional work and leisure time for billions of people around the globe struggling through their daily commute, but they also have the potential to save hundreds of thousands of lives by eliminating the human errors that cause nearly all traffic accidents. Unfortunately, there is a long road ahead for the technologies, legislation, and infrastructure that will be needed to make autonomous vehicles a part of peoples’ day-to-day lives.

Where Does AI Come In?

For autonomous vehicles to operate, they must essentially replace a human driver with their AI facsimile. To fulfill this goal, the autonomous vehicle needs to mirror the driver’s ability to assess a given situation, analyze available options, and take the best course of action.

Humans do this by perceiving the road with their eyes and ears, using their brains to analyze that input, and then physically engaging the steering and braking systems to implement the best course of action.

For an autonomous vehicle, the process takes several more steps and much, much more hardware and software. That is not to say there are not similarities; where a human has eyes, an autonomous vehicle would have various forms of sensors; where a human has a brain, the autonomous vehicle would have an AI-enabled computer constantly analyzing current conditions and planning courses of action. Finally, the human driver’s muscles would be replaced with a series of motors and actuators to actually control the vehicle. The engineering is challenging but doable; achieving public acceptance will require some persuasion, including a near-perfect performance and safety record.

“Look, No Hands” Driving
(Source: Wikimedia Commons)

Autonomous Trucks

Beyond any safety considerations, there are economic incentives driving the self-driving market, as AI systems can largely replace the long-haul truckers who move products and supplies from Maine to California. Unlike a human operator, an AI driver can function 24X7, expediting deliveries and lowering costs. The effect is similar to that of an aircraft autopilot which can guide a plane thousands of miles with little or no pilot intervention. But while a pilot (actually two, for backup) is needed for take-off and landing, and to counter an autopilot failure, autonomous trucks can be unmanned (once their reliability is established).

Kodiak Robotics Self-Driving Truck
(Source: Wikimedia Commons)

Autonomous Taxis

Modeshift, a firm that helps transit agencies modernize their fare and data collection infrastructure, reports that “Another innovative and exciting prospect of AI in Transportation is its application in drone taxis. It is estimated that AI-based drone taxis will be able to facilitate intra-city transportation to reduce the strain on existing urban infrastructure. This could provide a valuable solution to municipalities that are already under tremendous pressure as they struggle to meet the demands of growing populations when it comes to smart urban planning.”  

The company adds that pilotless helicopters can “combat carbon emissions, eliminate traffic congestion, as well as reduce the need for expensive infrastructure construction plans.” Helicopters or other “aerial vehicles” can fly remotely piloted or autonomously, delivering people and/or packages with incredible precision.³

Other Applications

In addition to enabling autonomous vehicles, AI will serve the Transportation sector by providing or enhancing the following.

Intelligent Traffic Management

Outlining the possibilities, analyst Jeremy Bowman says that “smart traffic lights [could] turn from red to green when there’s no traffic coming one way. AI sensors could also determine the best times to adjust traffic patterns for rush hour, for example, by making a two-way street one way.

“AI tools and algorithms could also predict congestion and improve traffic flow, thereby designing traffic systems that save people time, speed up the transportation of goods, and have beneficial [byproducts] like reducing pollution.

“Less-than-truckload (LTL) carriers … stand to be among the beneficiaries of intelligent traffic management as they would save time shipping goods and turn over their trailers faster.”⁴

Safe Driver Monitoring

AI can be instrumental in monitoring both cars and drivers. As Modeshift suggests, “By adding computer vision to car cabins, [it’s possible for AI] to predict a change in a person’s emotional state,” or detect “signs of drowsiness. These types of advanced driver assistance systems will ensure the safety of the driver, [any passengers], as well as other people on the road. They can promptly alert drivers whenever their abilities … are impaired due to fatigue. Additional alerts will [sound] for [any] signs of distraction, adding an extra layer of safety.”⁵

Flight Delay Predictions

According to Modeshift, “It’s estimated that AI-based technology will revolutionize the aviation sector by solving conventional problems such as flight delays. By leveraging and merging computer vision with data lake technology, aviation companies can put a greater emphasis on improving customers’ journey experience by reducing wait time and offering better quality service, at an improved, expedited pace.”⁶

Building & Repairing Infrastructure

The New York Times reports that “At a time when the federal allocation of billions of dollars toward infrastructure projects would help with only a fraction of the cost needed to repair or replace the nation’s aging bridges, tunnels, buildings, and roads, some engineers are looking to AI to help build more resilient projects for less money.

“In Pennsylvania, where 13 percent of the bridges have been classified as structurally deficient, engineers are using artificial intelligence to create lighter concrete blocks for new construction. Another project is using AI to develop a highway wall that can absorb noise from cars – and some of the greenhouse gas emissions that traffic releases as well.

“[AI] works by analyzing vast amounts of data and offering options that give humans better information, models, and alternatives for making decisions. It has the potential to be both more cost effective – one machine doing the work of dozens of engineers – and more creative in coming up with new approaches to familiar tasks.”⁷

The Future

To help ensure that AI is applied fully and fairly to today’s – and tomorrow’s – transportation systems and services, the US Department of Transportation has identified a broad spectrum of “real-world transportation applications using artificial intelligence.”⁸ 

Some of the more interesting initiatives include:

Providing pre-trip planning and en route travel information to travelers with disabilities

AI can be integrated with current travel planners to learn a user’s [preferences] including frequently visited locations to create a personalized trip itinerary based on their physical and cognitive abilities. This day-to-day itinerary can include hotel, shopping, sightseeing locations, and restaurants that are tailored to their unique preferences, abilities, and other user recommendations.

Detecting and predicting traffic incidents to respond efficiently and proactively

AI techniques such as machine learning and NLP can be used to detect and predict incidents based on data from sensors, videos, and images, … third-party weather and traffic information, social media data, and other data sources.

Using AI to detect potential incidents in video feeds or images could reduce the incident detection time. AI could potentially be embedded within CCTV cameras to detect incidents, or AI could be used in the [Traffic Management Center (TMC)] to process the CCTV feeds in real-time to detect incidents. AI can hasten the data analysis process and allow TMC operators to respond more efficiently and proactively to incidents.

Planning and maintaining road infrastructure from smartphone video data and AI

Mobile applications can use computer vision to process and label road video data collected from smart phones. For example, an AI-enabled application could automatically assess pavement conditions, traffic signs, road surface conditions, and weather conditions from anonymous user-submitted data. With this information, the application could alert crews to downed, damaged, or visually obscured signs. Additionally, using color grading, AI could derive air and road temperatures and road roughness information. Overall, applications like these could be useful tools for TMCs, road maintenance crews, and others in transportation planning and maintenance.

Increasing efficiency of tracking and monitoring fleet repair and maintenance

AI can improve fleet maintenance in a variety of ways. First, AI can reduce unplanned truck downtime by monitoring engine data in real time and alerting managers to abnormal metrics. Second, AI can increase efficiency in the repair process by providing technicians helpful insights that are difficult to see or track over time. Third, AI can improve fuel efficiency by detecting potential pressure problems before fault codes occur. Fourth, AI can help ease data overload by fusing data from multiple sources and highlighting the most critical areas.

Improving emergency planning by identifying high-risk crash locations, identifying populations vulnerable to natural calamities, and planning for evacuation needs of specific population groups

Agencies can improve their emergency and evacuation plans by identifying high-risk crash locations, identifying populations vulnerable to natural calamities, and being responsive to the needs of specific population groups, who require assistance during local or multi-jurisdictional emergency evacuation. These specific population groups include people with disabilities, people with medical conditions, the aging population, people with no access to transportation, and people with pets.

AI can be used to identify high-risk crash/incident locations that can hamper emergency evacuation operations under the threat of major natural catastrophes such as Hurricanes Sandy and Katrina. Machine learning techniques can use historical crash and incident data including occurrences during disaster conditions to produce predictions for similar future events. Machine learning can be used to predict high-risk facilities and localities, and potential severity of impact. These predictions can be used to identify vulnerable populations and develop specific plans for evacuating people who require assistance.

Web Links

Modeshift.com: https://www.modeshift.com/
US Department of Transportation: https://www.dot.gov/
US National Highway Traffic Safety Administration: https://www.nhtsa.gov/
US National Institute of Standards and Technology: https://www.nist.gov/

References

¹ Traffic Safety Facts Annual Report Portal. August 2023.

² “AI in Transportation Global Market Report 2023.” The Business Research Company. December 2023.

³ “What Is the Future of Artificial Intelligence AI in Transportation?” Modeshift. May 19, 2023.

⁴ Jeremy Bowman. “The Future of Logistics: How AI Is Changing the Transportation Industry.” The Motley Fool. June 2, 2023.

⁵ “What Is the Future of Artificial Intelligence AI in Transportation?” Modeshift. May 19, 2023.

⁶ Ibid.

⁷ Colbi Edmonds. “New Tool for Building and Fixing Roads and Bridges: Artificial Intelligence.” The New York Times. November 19, 2023.

⁸ “Identifying Real-World Transportation Applications Using Artificial Intelligence (AI).” US Department of Transportation. July 2020.