AI: MAKING SHARED MOBILITY SMARTER, by Madeleine Resener 

Booming computing power, big data and deep-learning technologies facilitate so much the development of artificial intelligence (AI) that we can expect to see a vastly different transportation landscape.

The transformation will include driverless buses routinely shuttling people safely to their destinations and smart, sustainable vehicles performing tasks such as ploughing snow, collecting garbage or delivering food and mail. Here are a few of the ways in which AI will enrich our daily commuting lives.

 

IMPROVING THE USE OF PUBLIC TRANSPORTATION FOR THE DISABLED

 

Californian industrial and public researchers are working on an artificial intelligence project on the Caltrain (the commuter rail line on the San Francisco Peninsula and Santa Clara Valley). The goal? To give passengers with vision, hearing, or other disabilities real-time information in order to help them find the right track, platform, and train, as well as the optimum spot for boarding. The technology could also alert passengers when to disembark and confirm that they are on the right train. The Internet of Things system taps into the cloud, smartphones, and data devices called “beacons” at Caltrain’s Diridon Station. The project could be expanded to include BART – Bay Area Rapid Transit System – and high-speed rail, or used for transit terminals nationwide.

 

SMART CHARGING FOR LOWER ENERGY CONSUMPTION

 

Until now, recharging the batteries of electric buses has been a challenge for the cities that use them. The system requires putting the buses on charge when they return to the depot and leaving them there until their departure the next day. This may provoke peaks in energy consumption that can wear down the batteries and even result in overcapacity fees. Artificial intelligence is about to change that. Smart charging involves smoothing out the consumption so that the peak is never reached and the battery life is preserved. Keolis is experimenting with this new method that will ensure the availability of energy for residents when cities fully electrify their bus fleet.

 

RESOLVING URBAN CONGESTION

 

Leveraging AI to understand and predict passenger movement is one of the latest innovations that aims to resolve global urban congestion problems. The software leverages low-power smartphone sensors to detect and predict passenger movement across various modes of transport. It then processes real-time sensor data from smartphones and wearable devices without the need for any other external hardware. Because the solution leverages fundamental AI techniques to model passenger movements across several modes of transport, it can identify key transition points such as passengers waiting at the platform to getting on the train. So it helps public transport operators to get an end-to-end understanding of journeys.

 

MAKING PUBLIC TRANSPORTATION SAFER

 

Experts have proposed an AI-based system to help make better decisions relating to safety so that riders can enjoy a more comfortable and secure journey. The idea is, first, to let an algorithm monitor all the incoming passenger communication via Twitter, Facebook, and online chats. The information will then help to determine whether they relate to a critical emergency situation, such as fire, crime, or faulty equipment. Once identified as an emergency, the system will decide which department and location are best equipped to handle the situation and automatically push the customer’s message to mobile phones of all the relevant stakeholders. It all makes for a safer, more enjoyable journey for passengers.

 

ANTICIPATING BREAKDOWNS IN ROLLING STOCK

 

Transit operators stand to gain big by using AI to carry out maintenance. Instead of relying on diagnostic tasks performed under human supervision, “predictive maintenance” anticipates breakdowns by detecting data inconsistencies. It relies on collecting and analysing information from the millions of sensors located on critical train components to anticipate maintenance requirements before accidents occur, thus significantly decreasing stoppage times and costs. Predictive maintenance also uses diagnostic, warranty, survey, federal and social data sources with advanced analytics to detect and predict which parts, or combination of parts, are potential issue points. The cost savings for rail operators in terms of avoiding major repairs or recalls is significant – as much as 25% by some estimates.

 

PROVIDING MORE EFFICIENT CAMERA SURVEILLANCE

 

Camera surveillance has become an integral part of protecting passengers. Analysing videos, however, isn’t easy: it requires time, attention to detail, and discernment by the security personnel. Keolis is experimenting with a way to make security cameras used in public transportation more effective. The solution involves connecting the cameras to a software system that can identify an abandoned bag in a precise location in just 10 seconds and retrace it to its owner.

 

SIMPLIFYING TRAVEL BY INTEGRATING VOCALBOTS IN PUBLIC TRANSPORT

 

Artificial intelligence-powered voice interfaces, or vocalbots, are useful to give passengers quick answers to questions like “When is the next bus to work?” or “How do I get to the dentist?” Using the transport network information already stored and a journey planner system, the solution is able to provide an answer within a few milliseconds. Today, the idea is to make it easier for customers to choose public transport by integrating a platform with a hands-free speaker you control with your voice. This could be particularly useful for elderly and disabled people who may find it difficult or even impossible to use smartphones, but still need to access all kinds of travel information.

 

 

More info to discover in Pulse #3 here: https://pulse-mag.com/en/index.html