November 9, 2017 | Written by: Hugh Palmer
Categorized: AI/Watson | Automotive
Share this post:
A few years ago, I had an experience traveling through the Washington, D.C. area with my father, who is in a wheelchair. He continually asked me about where we were and what was going on. I could sense that he was nervous about being in a wheelchair in an unfamiliar environment. I knew that I could also do something to provide my father and others like him with a new sense of security while traveling.
Designing a vehicle that cares
At Local Motors, we design, build, and sell vehicles and vehicle technology. We are an American vehicle manufacturing company focused on low volume production of open source vehicle designs, manufactured in micro-factories across the country. Local Motors uses co-creation and micro-manufacturing to develop and design vehicles for the local community, and build them in a local factory. Around the time of my trip to Washington, we were just starting to design Olli, the world’s first self-driving, cognitive vehicle.
The experience with my father made me think about how Olli should provide a new level of security to its passengers, taking care of them as a caregiver would. In the absence of a driver, the vehicle should attend to their needs, answer their questions and provide simple services. It’s something that I believe all of us could use.
Crowdsourcing: the key to designing #AccessibleOlli
At Local Motors, we develop our vehicle concepts through co-creation, which is the crowd sourcing of ideas and design concepts. We set the goal of crowdsourcing an accessible, self-driving vehicle using Internet of Things (IoT) and cognitive technologies. This was a combined initiative from Local Motors, IBM and the Consumer Technology Association (CTA) Foundation, whose mission is to link seniors and people with disabilities with technologies that enhance their lives.
We gave the public the challenge: the #AccessibleOlli project – a crowdsourced effort to create the world’s most accessible mode of transport for those with disabilities or impaired mobility.
#AccessibleOlli is a public vehicle. It provides a service to the public, so it needed to accommodate people of all ages, our elderly as well as our disabled. These disabilities include physical and cognitive impairments as well as speech and hearing difficulties. It’s our goal that #AccessbileOlli meets the needs of all passengers, whatever language they speak or challenges they face.
We debuted Olli in June 2016 at the Local Motors National Harbor. Today, Olli is entering its pilot phase, currently operating in Denmark and will be seen operating in the United States in early 2018. We’re excited to see Olli in many U. S. cities, as well as in large cities in Europe.
Training the world’s smartest bus driver
I would advise people who work with IBM Watson and cognitive technology to consider an experience from both the human and machine point of view. As a product manager, I often ask myself what a human would do in a particular situation. For example, what services would a driver provide? What would an experience be like if the machine were replaced by a human, and whether the same service can be delivered without the human? Running through the various scenarios helps me understand what cognitive technology can provide for users and for the public.
One challenge to creating self-driving vehicles that we can solve only with cognitive technology is to adapt the vehicle to the various environments in which it will drive. It’s daunting to think of training the vehicle through software algorithms to behave properly in so many uncertain environments.
A self-driving vehicle needs to have the intelligence of a bus driver, but when the technology first comes out it’s like an uneducated driver. It begins its experience on the roads and has to mature through years and years of experience, just as a human would. But no human can take the time to sit down with the vehicle for 12 years to teach it how to drive.
Machine learning allows Olli to learn rapidly; to learn on its own as it encounters new situations. It enables the vehicle to ask questions, to understand its environment and continually learn. Machine learning is essential for the rapid development and maturation of a self-driving vehicle.
That day with my father made me realize the importance of our self-driving vehicle and the level of security it could provide to its passengers. As I answered his questions, I realized how important this project was to someone like him, or to anyone who might need this kind of service. I am pleased with how far along we’ve come and think my dad would feel the same.
For more information about IBM solutions for the automotive industry visit here.
Learn more about #AccessibleOlli by watching the video interview below with Hugh Palmer: