Arizona-based startup Local Motors has unveiled Olli (earlier post), a 12-passenger minibus that may hit university campuses in Las Vegas and Miami later this year. In Helsinki, Finland, two driverless buses were employed last year—one of the first programs of its kind. Likewise, a plan in Western Australia aims to do away with timetables, supplementing their autonomous transit system with an on-demand bus service that provides service to elderly and disabled people who don’t live along standard-route bus services. Shuttles like these emulate systems currently used in the Netherlands.

These kinds of programs seek to do more than economize public transportation. They aim to change the way humans think about public transit, making it their first choice, not last, for getting around. Otherwise, smaller driverless vehicles (including taxis, rideshare services and private cars) could potentially eliminate the need for public transit. After all, why remain beholden to a fixed bus route when a car can pick you up and drop you off door to door?

The benefits of autonomous buses. From a mathematical perspective, buses take up less space, use fewer resources and accommodate a greater number of passengers. This alone holds tremendous offerings in mobility efficiency and reduced environmental impact, not to mention cost savings.

To get an idea of just what benefits driverless buses could bring, the International Transportation Forum (ITF) created an autonomous bus simulation based on Lisbon traffic to see how a European city might cope with an exclusively self-driving bus system. The results were staggering.

Study results, as summarized by Vox, found that, “Shared vehicles could make our cities dramatically more livable.” The research found that “traffic emissions were reduced by one-third, and 95 percent less space was required for public parking,” and the vehicle fleet necessary would only be 3 percent of today’s fleet. Passengers, for the most part, wouldn’t have to wait around for transfers, and spaces previously devoted to parking would make way for open, livable spaces.

The study does point out, however, that the benefits wouldn’t be felt until private cars on the road are minimized by at least 60 percent. ITF also noted that despite its inefficiencies, single-occupancy cars will dominate in three critical areas: flexibility, comfort and availability. Moreover, technological innovations have helped lower the overall costs of cars while disposable income in many countries has increased. Thus, the private car will likely continue to reign supreme.

So, how do we bridge the gap between the personalized travel experience offered by the private car, and the compromise of a fixed-route system of mass transit?

Tapping into the rideshare model. An emerging concept for public transit is to mimic rideshare services. Like algorithms used for Lyft and Uber, which sync carpoolers and drivers around optimized routes, so autonomous buses and taxis might offer similar operations but at greater capacity. To put it another way, the bus stop could be your front door.

However, city planners like Jarrett Walker explain that mass transit is already strategically planned along high-demand corridors, optimizing use of urban space and accommodating its ever-growing density. A rideshare-like system that adds more trips may not bring with it many benefits unless it can serve more people per trip.

As Walker notes, “Transit achieves high ridership by being useful and liberating to lots of people, not to any particular kind of person. High ridership arises from diversity, not specialization.” Therefore, it may be useful to consider developing certain routes that serve groups of people with similar interests or needs, or to find a compromise between fixed routes and rideshares.

The possibilities of microtransit systems. In 2015, the one-year pilot program, Bridj, created a “microtransit” hybrid of ride-hailing and bus service in Kansas City. Acting as a rideshare-style bus, the program guaranteed passengers a ride within a five-minute walk from a given pick-up point and destination. It wasn’t as fixed as public transit, or as exact as Uber or Lyft, but it created an as-needed model for those who weren’t within walking distance of a bus route or train station. According to Laura Bliss of The Atlantic’s CityLab, in her article “Bridj Is Dead, But Microtransit Isn’t,” this kind of hybrid system relies on “super-smart software that formulates routes and spits out pick-up spots in real time, based on demand, for any type of rider.”

While Bridj is now defunct, it has, perhaps, hit a nerve in the exploration of mobility efficiency that simultaneously upholds levels of flexibility and comfort traditionally associated with private cars—a feature that ITF maintains is a stepping stool to a full fleet of autonomous buses.

What comes next? China’s tech giant, Baidu, is trying to become to autonomous cars what Google’s Android has done for smartphones. The company has challenged other autonomous tech companies such as Nissan, Google and Tesla to make autonomous cars available to consumers by the year 2020, and has even open-sourced its technology. Andrew Ng, Baidu’s Chief Scientist, recently projected that the first fully autonomous vehicles we see on the roads will be buses, followed by passenger cars: “A large number of such vehicles to be in operation within three years (= early 2019) and mass production to be in full swing within five years (= 2021).”

But with excitement comes complexity. In an LA Times op-ed piece, Steven Greenhouse explains how driverless technology could ultimately cost 5 million people their jobs, which amounts to about 3 percent of the workforce. Not only that, “many households may decide they don’t need to own a car. As a result, many auto dealerships, car washes, gas stations and parking lots might shut down, causing additional layoffs.” Which is why Greenhouse argues for government intervention to “cushion the blow” so that we don’t see a continuation of the 5 million factory workers who’ve already been laid off since 2000.

There are also significant implications for auto insurance. In all states except New Hampshire, some form of liability coverage is required, and premiums are determined in large part by risk. Because autonomous driving tech is slated to alleviate risk at prodigious levels, we might see premiums go down. Not to mention, driverless tech relies on high-octane data exchange, whereby insurance companies can create remarkably accurate pricing profiles. Safe drivers, especially, can expect to benefit from this.

But high volumes of data being exchanged between cars and roadway infrastructure are likely to mean buses, trucks, and passenger cars will all become a sought-after data cache by both businesses and hackers. Traditional insurance companies will therefore need to rethink their models and product offerings as cyber security becomes a mounting issue. And, we may very well see a large uptick in other lines of financial coverage—namely, affirmative cyber insurance—to assure people’s sensitive information is adequately protected.

For now the impacts of autonomous vehicles on insurance remains a bit of an unknown. But it’s clear that as driving technology continues pushing forward, so too must insurance.

About the Author. Dave Crichton is a Vice President of Product at Esurance, where he is responsible for pricing, underwriting and third-party data integration. Dave has 22 years of experience in the industry, focused primarily on the underwriting, pricing and innovation of auto insurance products.