After 30,000 Driverless Lyft Rides Consumers Rate it Almost Perfect

After 30,000 real-world driverless Lyft test rides in Las Vegas consumers have rated it an amazing 4.95 out of 5, says Lyft’s Chief Strategy Officer Raj Kapoor. He says that in the last 12 months the system’s gotten smarter and the ride has gotten smoother. “It has measured reactions and acts like a really good driver versus maybe an inexperienced driver,” says Kappor. “That’s a big change.”

Raj Kapoor, Chief Strategy Officer at Lyft, was interviewed by PCMag at the Consumer Electronics Show in Las Vegas:

Consumers Rate Selfdriving Lyft Rides 4.95 Out of 5

What’s changed changed in the last 12 months with our selfdriving tests is first of all the system’s gotten smarter. The smartness comes out in terms of planning and prediction. You can now tell how smooth the selfdriving ride is. If it’s seeing pedestrians or lots of cars it doesn’t make knee-jerk reactions. It has measured reactions and acts like a really good driver versus maybe an inexperienced driver. That’s a big change.

Two is that we broadened the area that we’re operating significantly. We’re operating a geofence that covers almost all the major hotels in Las Vegas and you can go anywhere in that area versus very specific points.

It’s one of those rare instances where a cool demo from CES right away becomes a live service. We have now 30,000 rides that we’ve had in the system and so far the feedback has been awesome. Consumers have rated it a 4.95 out of 5 stars and 9 out of 10 people that go on a ride would come back and do it again.

We’re really quite pleased with it. I think people have a lot of questions around self-driving, there’s some fear, but once they get in and do the ride they are really excited about it and ready to do it again.

It’s using the same scale as normal Lyft rides and in their mind it’s that same perception of how comfortable, clean, did the person or the robot drive well, all those things go into account. The automated Lyft is rating pretty high but the human drivers definitely get up there too. But 4.95 out of 5 is very respectable.

Lyft Using Las Vegas Data to Perfect the Selfdriving Ride

We can notice a lot of things with this real selfdriving test. Las Vegas is a great proving ground because there are so many people from around the world that come here. You’re not just having residents that are here. You’re having people that are using it for their vacation, using it for going back and forth, so we’re able to collect lots of data from a big diverse group of people.

We’re able to see what the repeat use is like. What do they like about it? What do they dislike about it? How much do they like to walk to the vehicle versus not? There are all these little nuances that go in.

Another example is around remote assistance. We noticed that the people love to have a conversation with the safety driver. They’re so excited at that moment. The question we have is as we move toward a future without a safety driver is how do we still get that interaction with the consumer? Can we have a remote assistant do that? We’re learning all these little things by being in the real life out there with people.

In fact, there are two people in the front. There’s a safety driver and also a safety engineer that’s in the front. So usually it’s the safety engineer that’s answering the questions. The safety driver is very focused on the road.

What doing now is developing remote assistance systems where you can talk to a lift operator and ask the questions that you need to ask via a very conversational interface with the consumer. There’s a lot going on there and there’s a lot going through their head around building trust. What is this car seeing? Is it acting the way that I would act? Then asking questions about how does that technology work? How can it do this?

More Human Drivers Will be Needed, Not Less

Going to the question around drivers. We still believe that there will be a need for more human drivers than there are even today. If you look at it these ride-sharing services in the US make up about 0.5 percent of vehicle miles traveled. Even if we go to 5 percent and the majority of rides become self-driving, if you do the math on that growth of the overall number of rides, you need even more drivers than you have today.

Also, this technology is great but it’s going to be slowly rolled out. It’s going to take a while for it to be able to do all conditions, all places, all the time.

We haven’t even begun to imagine around the new economy that comes out of self driving vehicles. There’s so much infrastructure that needs to be built around parking and charging and even mundane things like cleaning cars. Then there are serve there are groups of people that need to have assistance, whether it’s elderly people, people requiring physical assistance, or young children. We think there are going to be opportunities abound.

Obstacles to Overcome with Driverless Cars

There are a number of obstacles to overcome with driverless cars. Can I trust that this vehicle is going to operate safely, not just for me but for all the environment and community around us?

Two, is the technology itself. I think we’re on a good path and we’re improving but it’s going to take some time to get there. The other piece of it is that the cost of the vehicles are significantly expensive right now. They’re using expensive computers, expensive sensors, and they’re not made necessarily to last for a long time because they’re in a lot of R&D stage.

So the reliability and the costs have to get better. Then on the government side, the regulations that need to allow for this to flourish. We’re seeing good progress there. If we continue to have a federal level on safety standards then that’s something that’s really positive because it’s not going to be that you have to create a specific car for specific jurisdiction. We think those are the barriers but they all seem very doable.

Surprised at How Quickly Micromobility Has Taken Off

We classify all this (scooters and bike) as micromobility. What we found is that there’s really an unmet need for that zero to two-mile range short distance trips. Yes you could take a Lyft or you could potentially walk but you’re in that zone, especially if it’s a half a mile or more. It is a really convenient thing to do. It’s also a really fun thing to do whether you’re biking or scooting and especially if it’s electric propulsion.

I’m also surprised at how quickly that has penetrated but I think we’re living in a world now where there’s mass adoption, there social networks, and the innovation that’s coming that came in software so fast we’re seeing in hardware. New versions of scooters are about every month. There’s some loss rate and breakage rate that is acceptable given the high usage and it works for the consumer because it’s still a very reasonable price to get around town.

Lyft Helping Society Shift Away From Owning a Car

For people not to own a car, it’s really going to be a stitching together of a number of different modalities. Whether it’s bike, scooters, walking, or ride-sharing together, to provide a really good alternative to owning a car, which is expensive, a hassle, has parking, congestion, and emissions. It’s a really big problem that all these things together are going to be solving.

We want to get them from point A to point B in the most convenient way that they could get there and without owning a car. That’s the key criteria.

I think there will clearly be a big group of the population, especially that lives in cities, that really will not be able to justify owning a car. All the use cases that you’re thinking about you could utilize by short-term rentals, car sharing, ride-sharing, micromobility, or also public transit.

We’ve integrated into public transit and it’s the best way to get around in a lot of instances and we’re feeding into that. There’s a lot of inertia still about buying a car and some people are still wrapping up a car in their self-worth and their identity. That’s changing, especially with young people.