The electric, 3D-printed, self-driving Olli minibus was giving test drives at the IBM World of Watson conference in Las Vegas last month, while passengers asked the onboard Watson artificial intelligence platform for local restaurant recommendations.

IBM is betting the company’s future on Watson-based cognitive technology and cloud computing for enterprise clients, and it’s promoting the bejesus out of Watson across all media channels. The recent 60 Minutes story on the supercomputer, new partnerships with General Motors and Slack, and the slick Watson ads with celebrities like Bob Dylan and Serena Williams, are all driving IBM’s messaging storm to make IBM machine learning a household conversation.

The company has been slow to push cognitive technology to travel consumers in any gripping way, however, except perhaps for the development of Hilton Worldwide’s robot concierge, Connie. IBM’s travel division is presently much more focused on improving efficiencies in operations for airlines, railways, and shipping companies, keeping Watson mostly out of view for travelers.

Partnerships with attention-grabbing travel technology platforms such as Olli could shift that.

Local Motors in Phoenix designed the 12-seat autonomous shuttle bus to look like a lovable soap bar on wheels in an attempt to make it more modern, more fun, and less intimidating to riders of all ages. Olli is already being tested in National Harbor, Maryland, outside Washington, D.C., with the aim of expanding transportation options to the new MGM National Harbor hotel opening next month. Permits are also in place to launch Olli in Miami and Las Vegas in 2017.

All three of those are resort cities with neighborhoods that have a high demand for public transportation in small and easily navigable areas with low speed limits and limited parking.

Mass transportation usage on a large scale is likely a distant scenario. According to this Inverse story, Local Motors sees Olli initially rolling out mostly in places like college and corporate campuses while Watson “learns” how to be a better driver. And because Watson is a cloud platform, everything that one Olli learns, every Olli learns.

“Olli is a robot you ride in, and it’s going to give mobility more mojo,” Joe Speed, head of the Watson IoT AutoLAB, told Skift at IBM WOW. “This is co-created industrial design from Local Motors, which crowdsources solutions to engineering problems. We were inspired by this, and now their developers and our developers are working together to co-create the future of cognitive mobility.”

Olli’s top speed is 30 miles per hour, although it normally operates between 10-15 miles per hour, with 40 miles of range on a single charge. It’s best suited for what the transportation industry calls a “first-last mile” solution, meaning distances that are too long to walk and too short to be worth the drive, especially when parking is a challenge.

Navigation is determined through a combination of GPS and lidar technology, which is like radar but uses laser instead of radio waves, and Speed said that all Olli journeys are supervised by remote live operators.

The user experience is designed to eventually look something like a cross between UberPOOL and any typical hop-on/hop-off transit bus. Riders will use an Olli app to signal their request for a shared ride from point A to B. After pickup, the Watson brain is continually determining the best route in real time to get the rider and as many as 11 other passengers to their various destinations as efficiently as possible, taking both climate and traffic conditions into consideration.

Meanwhile, riders can ask Watson questions about local special events, dining options and Yelp reviews, tourist attractions, hotels and room sharing, or anything else. Looking ahead, Watson aims to personalize answers on a deeper, more customized level based on users’ social profiles connected to the Olli app.

“The UX-HMI (user experience/human-machine interface) stitches together a lot of things, like Weather Company APIs, which IBM owns, and Yelp Fusion APIs,” said Speed. “As the same way Olli is crowdsourced local mobility, Yelp is an amazing source of local knowledge, and we use Watson to put those together.”

Speed also emphasized that the shuttle’s design is paramount from a public engagement perspective, resembling somewhat of a physical manifestation of Watson’s friendly and lovable geek voice. That alignment, he said, supports the IBM and Local Motors collaboration as much as the hand-and-glove combination of self-driving hardware and AI software.

“We’ve had people coming up and hugging Olli,” Speed explained. “It’s like, you see sports cars modeled aggressively, almost like a predator sometimes. This is the industrial design equivalent of Hello Kitty.”

Reimagining The Flight Attendant Experience

Heidi Fillmore, global industry executive partner, IBM+Apple, travel & transportation, at IBM also attended the conference. Part of the Apple/IBM partnership, officially known as IBM MobileFirst for iOS, Fillmore is working to develop better applications for end users in the travel and transportation industry.

She said the mission for MobileFirst is to use the Apple OS and user interface to create more intuitive and user-friendly apps. In the past, enterprise clients have often slapped a front-end onto an existing IT system to build new apps, without regard for how employees are going to actually use it in the workplace.

“We just had Finnair here saying they’re taking the bold step to build apps their employees don’t dislike,” Fillmore shared. “So we look at it holistically. Applications need to be intuitive, and they shouldn’t require a month of training.”

One of her division’s newest projects is the Passenger+ iPad/iPhone app for flight attendants, presently being used by Scandinavian Airlines and Air Canada. The app shows all passengers on any given flight who are missing connections, have lost their bags, or have special service requests. Basically, it makes it easier for attendants to approach flyers on a one-to-one, first-name basis whenever there’s an issue of any kind.

The app’s detailed seat map also identifies VIP passengers flying in coach.

“Airlines like that attendants can see top-tier members seated in the economy section, because they’re always asking, ‘How do we make those people feel special?'” said Fillmore. “Attendants are hurried and scattered at the best of times, but now it’s a lot easier to go give the guy in 12C a drink, because his information is immediately accessible.”

The Passenger+ app was also designed to allow attendants to rebook flights mid-air for passengers who missed their connection, but Fillmore said that U.S. and European labor laws are causing some challenges. Attendants can however reroute bags on the fly.

Presently, Passenger+ is not integrated with Watson, but that’s going to be the mandate for 2107. The goal is for attendants to have deeper profile information on passengers to better personalize the individual flight experience.

“Right now, we can filter passengers by all different things, like tier status, but we want to go way beyond that with Watson,” she explained. “So things like: Where are my most important customers with the biggest social footprints? And how can cognitive to do more targeted selling, and dynamically package content for the end user?”

So why isn’t IBM isn’t blowing out Watson more in the travel space from a consumer perspective, like the company did with Hilton’s Connie platform? At the IBM World of Watson conference this year, there was very little show space dedicated to travel, or transportation, for that matter.

“A lot of our Watson initiatives in travel aren’t as sexy because they’re more on the operations side, in terms of predictive analysis, managing inventory, or discretionary fuel planning,” Fillmore said. “I think we in the travel and transportation industry are doing a lot of stuff, but we didn’t maybe package it very well for this conference. So I think we’ll definitely showcase T&T (Travel & Transportation) a lot more next year.”

IBM World of Watson 2017 should have Olli right up front and center, or a fleet of them talking to each other. The boxy bus was definitely a hit at this year’s show.

Or better yet, Olli and Las Vegas might be ready to shuttle attendees between their hotels and conference center. Seems a little fantastical now, but that would definitely drive some attention toward the future of cognitive travel.

Photo Credit: The photogenic, driverless, 3D-printed Olli minibus was a crowd favorite at IBM World of Watson 2016. Skift