Artificial intelligence to help improve bicycle safety and mobility in Santa Barbara County

SANTA MARIA, Calif. - A significant state grant for Santa Barbara County is intended to help pave the way for improved cycling safety and mobility through the use of artificial intelligence (AI).

"We're very grateful to the Department of Transportation, Caltrans, for providing us with $480,000 to basically transform the way bike maps are done," said Peter Williamson, a planner with Santa Barbara County Association of Governments (SBCAG). "We will pull together the public in public workshops, etc., to develop a model for artificial intelligence."

The funding will allow a collaboration between SBCAG, UCSB and Simon Fraser University in British Columbia, Canada to take shape and train AI to map out the entire bicycle infrastructure with Santa Barbara County.

"It will create an AI tool that will allow us to map bike comfort for all the routes that people might want to bike on in Santa Barbara County," said Trisalyn Nelson, UCSB Geography Chair. "You can imagine the number one barrier to getting more people on their bikes is concern about safety. People need safe places to ride and this map will help them make decisions about what is safe for them and their families."

Once completed, the new mapping system will replace an outdated model and give cyclists a new useful state-of-the-art tool that will enhance cyclist safety and mobility.

"The last bike map that was done for the county was done in 2013, so it's desperately in need of an update," said Williamson. "We have an opportunity to partner with UC Santa Barbara and Simon Fraser and create, a model that's sustainable, accurate and reliable in terms of mapping. It's something that's outdated and needs to be updated. Ultimately, what does this do for the community? It helps you identify a system that's comfortable to you and that makes it more safe to bike, and also makes it more easy to know where is the best place to bike."

According to SBCAG, the mapping project’s has four key objectives, which include:

• AI Training: Data from Google Street View, OpenStreetMap, and an advisory committee will inform, train, and develop a tested AI model.

• Infrastructure Classification: A diverse selection of individuals will consistently rank bike routes from easy to expert for the AI model.

• Mapping: A specialist in cartography will utilize the data generated by the AI model to create a digital and hard-copy bike map.

• Wayfinding Plan: Develop a universal, adaptable regional wayfinding system based on AI findings. This plan could help secure future funding for installing uniform wayfinding signs countywide, enabling cyclists to navigate easily with or without smart devices.

According to Nelson, she predicts that the earliest results from the new mapping project will be accessible in 18 months at the earliest.

"If you're a cyclist, you should be excited, but you should be even more excited if you're someone who wishes to be a cyclist, but struggles to figure out how to make that leap," said Nelson. "This project is really for you. It's to help you identify what are the safest routes and what are the most comfortable routes to get you started."