AI might be ready but are we?

In a time when technology seems to evolve on a daily basis, the Federal Highway Administration is looking to lay the groundwork on integrating artificial intelligence into the transportation network.

The report goes into great depth on the past, present and future of AI, and how emerging machine learning is expected to help create automated transportation systems.

“Machine learning applications offer the potential to supplant human work in a variety of” transportation management systems, the report says, including data analysis and automated “traffic imagery analysis, incident detection, traffic control and traffic signal timing.”

The report notes that AI technology has hurdles to clear before it can reach its potential.

“AI applications should find their way from research experiments and pilot demonstrations to fully scalable applications in the near term,” the report says.

Eventually, “these technologies are likely to come closer and closer to ‘plug and play,’ but currently there is still a reasonably large barrier between the dreams of AI-enabled (transportation management) applications and the need for significant expertise and investment to make those dreams a reality.”

The report says there is a clamor for AI technology in business and government markets, with much of the technology coming from Google, Amazon, Microsoft and Facebook.

Despite the barriers for actual use of AI, the report gives examples of how technology already is being integrated into transportation systems.

Several states (Nevada, Florida and Iowa) are using what is called “neural network technologies” to detect crashes and other incidents by employing “video image analysis and traffic prediction.” In another locality, Alexa is being integrated with a 511 system, and in another Google Assistant is being used to manage arterial traffic.

The federal highways report adds that driverless vehicles are likely to be in use in the near term and automated drones in the “medium term.”

There appear to be significant barriers to self-driving cars. The occasional high-profile Tesla crashes attest to that.

But recent news coverage also points out obstacles for such high-tech transportation.

The Detroit Free Press reported that U.S. lawmakers are battling lawyers groups over how to handle the civil lawsuit side of things—who can they target when an automated car causes a crash?

Another report, by the website Naked Security, cited issues with data used to train machine learning systems, which lack hundreds of bicycle and pedestrian images.

And a report last week by CQ Roll Call, indicated legislation may be close on federal, state and local laws and rules, but also noted significant issues yet to be resolved.

The article also noted there may be more urgency to address technology now, seeing that federal transportation authorities just allowed a California company, Nuro, to deploy thousands of driverless delivery vehicles.

Source: www.fredericksburg.com