Self-driving Olli shuttle with IBM Watson debuts in Washington area
- 18 June, 2016 05:51
Olli, a self-driving shuttle for 12 passengers running IBM Watson Internet of Things technology, made its debut in a shopping area of the Washington suburbs on Thursday.
While some "fine-tuning" of the self-driving features are needed, passengers, by this fall, should be able to ride around and speak directions to Olli on the private roads at the National Harbor shopping and entertainment area on the Maryland side of the Potomac River, according to a spokeswoman for Local Motors, the designer of Olli.
The vision is that Olli will be used in all kinds of venues, such as crowded urban areas, college and corporate campuses and theme parks. It could also become the "last mile" connection from a subway or bus stop to a job site. Miami-Dade County has ordered two of the vehicles for a pilot project there, said the Local Motors spokeswoman, Jacqueline Keidel.
Olli didn't give any rides to reporters and bystanders at its Thursday debut, but the vehicle dropped off Local Motors CEO John Rogers with engineers standing by to offer assistance if needed.
"Olli offers a smart, safe and sustainable transportation solution that is long overdue," Rogers said in a statement, adding that Olli with Watson "acts as our entry into the world of self-driving vehicles."
Olli is the first vehicle to use cloud-based cognitive computing from IBM Watson Internet of Things to analyze and learn from 30 sensors embedded in the vehicle. Four Watson developer APIs were used that allow Olli to interact with passengers: speech to text, natural language classifier, entity extraction and text to speech.
"You can say, 'Olli, turn right,' and if there's an obstruction, it will wait," Keidel said. "If a dog runs across the road, it will use sensors to stop." But Watson can also take broader questions like, "Olli, can you take me downtown?" or "Are we there yet?"
Since Watson is web-enabled, Olli will also be able to answer questions about popular nearby restaurants or historical sites, at least according to how Local Motors and IBM have described the vehicle's capabilities.
Harriett Green, general manager of IBM Watson Internet of Things, Commerce, wrote a blog to describe the importance of how users interact with driverless vehicles. "This user experience is the key to making self-driving vehicles a real part of our lives rather than a tech vision of the future," she said. "Olli is a remarkable example of the endless possibilities cognitive computing provides the transportation industry."
Green said IBM will expand its Watson IBM research by helping develop and create additional Ollis at Local Motors headquarters near Phoenix and at IBM Watson IoT's AutoLab, an incubator for cognitive mobility applications. "We have a long term vision with Watson," Keidel added.
Local Motors didn't divulge the cost of building the first Olli or what it will cost in production. The company has built a second Olli, which will be taken to Berlin for a demonstration, according to another company spokesman, Adam Kress.
Driverless shuttles and taxis are being developed in other parts of the world, including in Singapore, where a startup named NuTonomy recently gave reporters the chance to ride in one of its driverless electric cars. A tourist spot called Gardens by the Bay in that city is also using a driverless shuttle called the Auto Rider to transport visitors over its grounds. That project was partially funded by the Singapore government.
Back in the U.S., Local Motors has already earned a reputation for creating the world's first 3D-printed cars, including the Strati, and also incorporated 3D printing for Olli. About 25% of Olli is made of 3D-printed materials, while another 25% is made of parts from 3D-printed molds, Keidel said.
IBM has recently doubled-down on its Internet of Things capabilities. In addition to Olli, IBM promoted via Twitter an upcoming conference to "explore the future of connected buildings."
IBM is also collaborating with Cisco to use Watson technology inside Cisco routers to interpret data from sensors at the edge of networks. That approach will prevent too much unneeded IoT data from being stored in the cloud, to increase efficiency.