1. Why is it so difficult to launch autopilot cars? Waymo joint CEO Ma watts is emphasizing the importance of ensuring safety and continuing learning in the process of exploration.
2. Waymo’s self driving test vehicle has traveled 32 million kilometers on public roads, spanning 25 cities, with a simulated mileage of more than 32 billion kilometers.
3. Waymo driver has been tested under rainfall, high temperature, snow and other conditions to help deal with the situation that waymo does not have operation services at present.
4. Mavakana admits that people’s trust in autonomous driving technology is beyond their ability.
5. When using waymo one, people can play their favorite music because it is their space and they have privacy protection.
Tencent science and technology news on October 6th, Google’s parent driving company Alphabet Waymo is developing autopilot taxis, which is a significant breakthrough. Car Hailing applications such as online car Hailing giants Uber and LYFT have reshaped the city, allowing people to give up their cars, and usually more directly connect the buttons you press on your mobile phone with the real things happening in the world than any other application. Just this week, the transaction department of Bank of America said in a research report that Uber is the most revolutionary thing that has happened on mobile phones in the past decade.
But simply letting the car appear where you want it to appear is not the ultimate goal. It should allow the self driving car to appear in front of your house. Autopilot is expected to transform city and suburban area more thoroughly, reduce total vehicle ownership and allow commuters to spend several hours in travel time. This is a great dream. In the past year, we have discussed this topic with the CEOs of many technology and automobile companies. They are all studying automatic driving technology, but no one is really close to the above goal.
But there’s waymo. Waymo has quietly tested autopilot cars in Phoenix for several years, and they are about to launch a new pilot project in San Francisco. These vehicles are basically Chrysler Pacifica and Jaguar i-pace SUV, which have been greatly modified. They can drive automatically. Many waymo ride services are carried out without anyone else in the car, including no safe driver, only you and the robot.
Tekedra mawakana, CO CEO of waymo, recently received an exclusive interview to discuss the expansion of its automatic driving service to other cities, the current obstacles and other issues, as well as how she thinks the company will make profits over time. In addition, she discussed the regulatory problems facing the industry in attempting to introduce more widely automated cars.
The following is the full text of an exclusive interview with mawakana:
Q: How did it take long time to launch the autopilot car?
Mavakana: we’ve been studying autopilot technology for a long time, but I want to say that this is an engineering challenge for our generation, and that’s why it took so long. To realize automatic driving, I think it means two things: first, safety is the core of everything we do, and ensuring safety takes time. Second, we actually need to learn in this process, which is like a discovery process.
What’s interesting to me is when I think back to when I joined waymo nearly five years ago. I remember thinking, “the product and the market have reached the best fit point, and it is ready for commercialization!” but then I realized that this is not the case. You actually have to put technology into the real world, which is the way to help it get ready. This is what we always focus on doing, what we need to learn, and get feedback from Phoenix passengers.
Q: I’m curious. How’s your feedback? But first, you are the first co CEO, and your term of office only began in April. Setting up co CEOs was once an unusual arrangement, but now it has become a fashionable trend. Do you have any differences with another co CEO?
Mavakana: my colleague Dmitri dolkov is one of the founders of waymo. He has been committed to solving the problem of automatic driving throughout his career. Before taking the position of CO CEO, he was the chief technology officer and I was the chief operating officer. Therefore, we still focus on the two areas we know best. I think about how to commercialize, how to deploy technology and how to introduce it into the community. He focuses on developing waymo driver. Then let’s find out the best strategy: how do you rank them? What is the promotion plan? So far, everything has gone well. We have distinct but complementary skill sets.
Q: what can this technology do, and then what are you allowed to do with it, seems to face great challenges at present. I’ve asked many executives this question: How did you make a decision? What is your strategy for having to make some complex decisions?
Mavakana: This is something I had to relearn while working in waymo. I like to use data to make decisions. I think when you introduce new technologies, new business models or introduce consumers to experiences you have never had before, the real important thing is to know when you have enough data and adapt to your intuition. So I want to say that what people can experience in waymo is a strengthened sense of self-guidance. This is what I often use now because it is groundbreaking. There is no tactical manual, and no one has done it before. We think while exploring, which requires humility and openness in the learning process. So I need to learn and listen, and then make decisions. If this is not the right decision, we will turn and move on. So to have such a mentality, this is a learning journey.
Q: you can make people sit in autopilot cars, go everywhere and hitch a ride. I’d like to talk about this technology first, and then what kind of business you plan to use it to build. Let’s start with technology. Is waymo driver ready? Can you deploy and get a lift right away? The bottleneck is regulation and security. Are you still trying to find a solution?
Mawakana: just as we were chatting, waymo driver had appeared in the world. We have launched a service called waymo one in Phoenix, Arizona. Anyone who goes there can download the app and call a car to take them from point a to point B. When the car appeared, there was no one inside, including no driver. The beauty of this service in Phoenix, Arizona is that people can have their own space. They get into the car and can hold a conference call and do whatever they want. So this is happening. What I want to say is: the future has come. In about 130 square kilometers of land, which is equivalent to the area of San Francisco, so I don’t think this area is very small. This is our particular jurisdiction, but it is a safe practice. In addition, we have launched the trusted tester program in San Francisco. So now San Francisco residents have the opportunity to experience our service.
Q: do you think the core technology development has been completed and can be deployed on a large scale? Or are you still learning and testing the system?
Mavakana: I think there is a very different place in developing automatic driving technology, that is, it is not in one step, nor can it be deployed anywhere after the technology is developed to a certain extent. Geographically, with different street lights and different weather, each place has its own uniqueness. When we operate in two completely different environments, you will see what we have done and what you are doing now, which gives us the opportunity to learn from the technology stack. We have automatically driven more than 32 million kilometers on public roads across 25 cities. More than 32 billion kilometers of simulation, which is quite important. This is the way waymo driver learns. All learning results are shared in the team.
Q: in the area of Phoenix, I admit it’s quite large. However, the boundary of the area has not been expanded within Phoenix for a long time. You’re starting to launch services in San Francisco. Will we see faster market expansion?
Mavakana: I think the question is, what is the goal?
Q: I just press a button and the autopilot will pick me up.
Mavakana: Yes, that’s the customer’s goal. Our goal in Phoenix is to learn. The mall has 7 to 12 entrances. Which entrance do you want to get on the bus at? The reality is that when you give us the address, we may not be at the entrance you said, and no one in the car answers your phone or calls you. So what are the operational challenges associated with actually having fully autonomous services? This is what we have been studying. We are not in a hurry to expand. We are always listening to the feedback of passengers in order to make this service more and more pleasant. That’s what we did in San Francisco. Now, we have passengers, some of whom have lived in this city for a long time, and they are happy to provide feedback and bring this technology to market.
Q: the 32 million km you mentioned is waymo’s favorite data. You drive more miles than any other company, but the problem you describe has great limitations. It’s very small. You can drive hundreds of millions of kilometers, but you may still not know where the exit of the shopping center is. Is this your challenge? Is there any difference in your solution?
Mavakana: Yes, so when we talk about the mileage we have automatically driven, this is the data of all 25 cities, which is very important. We have tested rainfall, the high temperature of the Sun Valley, snow and so on. These miles are training waymo driver to cope with our current lack of operational services. When I talk about getting on and off, or the chaos in the parking lot, these are the specific contents of the services we provide and the feedback we get.
Q: there are many autonomous driving companies involved in all parts of the stack, some of which are combined with the depth of the car. Some companies are far away from this field. They are only selling sensors. Your team is mainly modified Chrysler Pacificas, and now i-pace is added. Do you think once the car can drive automatically, you will face the challenge of how to change the car itself? Or would you like to leave your car to someone else?
Mavakana: we are really happy to work with original equipment manufacturers (OEMs) to produce cars. As you said, we have cooperated with stellantis and Jaguar Land Rover. They have made cars for more than 100 years. In fact, we think that figuring out how to make waymo driver safer is a very exciting challenge, which is our concern.
Q: are you considering the actual wheels, transmission system, collision impact and so on, or is it a commodity for you? Can you change those in and out?
Mavakana: I think the underlying foundation itself is an evolving platform. Now we see all these innovations around electric vehicles. Although the internal combustion engine has been trying to improve efficiency for 100 years, the efficiency of gasoline engine is only about 25% and that of diesel engine is about 33%. Now we have electric cars, and their efficiency is more than 95%. This is an innovative platform, but it is not our concern.
I would like to say that we mainly focus on innovation in three aspects: the first is to build waymo driver. Can you really remove people and let the car drive automatically? Yes, we are doing it. It’s very exciting. Second, once you can do this, what business model can you unlock? Therefore, our focus today is to deploy waymo driver in three main business areas, including car Hailing service waymo one, long-distance truck transportation service waymo via and local delivery service. Finally, we will also focus on personal ownership of cars, which is the time for you to get the technology license.
But the third part of this innovation is actually a big challenge, because 40000 people die on our way every year. The real decision is that if 94% of these deaths are caused by human errors, we have the opportunity to innovate and remove humans from the system. This means you don’t need to rely on humans to be alert at any time. You don’t need a driver’s license. As long as you get on the bus, waymo driver can take you to your destination safely.
Q: Tesla has a software called “full automatic driving” (FSD). On this complex scale of automatic driving that no one understands, it belongs to L2 level system. You are already L4 level and want to develop L5 level. Do you think fully automatic driving, driver assisted driving system and so on are despised? Do you think this will spoil the good atmosphere of autopilot dialogue?
Marva Khanna: around 2013, Google’s autopilot project (before stripping Waymo) focused on driver assisted driving technology. What we have learned from experience is that people trust this technology beyond its capabilities. We’ve seen employees shave, sleep and plug the power into the back seat when starting this system, even though we told them it was beyond the ability of the system. So I think the problem is that we are human beings. When we got on the bus, we had other things to do. We have to leave home to go to work, but we have three emails to send, but the cat ran out. In short, there are a lot of complicated things to deal with.
So I think the problem is that you have to give people the opportunity to do what they want to do, rather than focus on driving. So the reason we focus on level L4 is that we believe this is the only way to safely deploy this technology in a way that improves road safety. This is our mission to survive. When you talk about driver assistance, it’s a very different business model. I have this kind of system in my car. It keeps me in the driveway. It makes it easier for me to drive, but it doesn’t let me ignore it.
Q: you just mentioned that one of the business models you will pursue in the future is licensed technology. Is this your goal of actively working with carmakers? Buy waymo’s stack, put it in your car, remove the steering wheel and sell the car?
Mavakana: Yes, we have announced our partnership in these areas, but we have always been open to it.
Q: are you competitive in that market? You have some partners, but do other companies focus on cooperation? Did you win those deals, or did everyone join automatically?
Mavakana: I think unless people create L4 level systems, it is difficult to start negotiating viable deals on L5 level capabilities. I can’t say what people may be doing, but I know no one else today provides L4 public services.
Q: will you authorize L4 level system? With Chrysler Pacificas and Jaguar i-paces, why can’t I buy a modified Chrysler Pacifica with automatic driving today?
Mavakana: we won’t authorize such systems. First, as I said, we are learning. Second, we operate it as a fleet without licensing the technology. Third, there is a map. You said yourself that this is a special jurisdiction. I think the automobile manufacturers will tell us that this is not a really feasible product for them.
Q: let’s talk about the actual business, first waymo one, and then truck transportation. Is this business mature? Is the income model set? Do you know how it will expand? What have you learned from operating it on your current scale and what needs to be changed with your expansion?
Mavakana: I think one of the reasons we did the early ride test in Phoenix is that we were actually learning to see what people care about? Everyone is thinking about how to build a sustainable city; What will the future city look like? How will electric cars help shape such a city? When you consider the pros and cons, we want to transport on demand. The reality is that this will lead to more congestion. So, will people make a trade-off? If this means that there is no idling and congestion in my city, will people say that I need another 8 minutes to get the car? Therefore, we have a lot to learn about these business models, which is what we are testing. I won’t sit here and say that we have figured out all kinds of commercialization opportunities in the next 20 years. We are actually listening to see how people will evaluate the service accordingly. This is testing pricing, all normal corporate strategies, but really, customers are always the core.
Q: in the data you have collected in Phoenix so far, what are you most surprised about the needs of customers and what changes you don’t want to make?
Mavakana: I think one of the most amazing things for me is that many people now feel excluded from transportation. I will give two different examples: one is those who have to take back their car keys because they are old. Second, we cooperate with the blind children’s foundation. These children’s 16th birthday is different from others. Therefore, it is a pleasant surprise to give people independent space so that they don’t need their parents, girlfriends or boyfriends to take them everywhere.
Q: do you have to adjust the price? Are people willing to pay the price you think they will pay?
Mavakana: people are willing to pay. We have a lot of demand. We won’t use the whole team for customer service. We are still using some vehicles to promote the progress of waymo driver, so our demand exceeds our supply in the market.
Q: how is your service different from all other existing self driving taxi services?
Mavakana: value proposition, right? As consumers, we will choose the best experience we think. Although you can enter the car Hailing service and take over their music, people usually don’t. The proportion of people who want to do so is very low. I just got into your car. I sat in the back seat, you sat in the front seat, you were driving me. I would only stay in your car for three minutes. When people enter waymo one, they will immediately start playing their own music. Because this is their space, they have this privacy. They can make a phone call without hiding.
Q: but in the final analysis, if this technology can be sold from you or other suppliers, Uber will own it and LYFT will own it. What made them choose you? What makes your price higher than others?
Mavakana: I think one of the premises of your problem is that everyone will have it, because everyone will solve the problem. Everyone will have it, because several companies will solve this problem, and then these companies will become our partners.
Q: now, when you open the LYFT app, you can call Toyota Camry anywhere in the world. It’s obviously surprising when this experience turns into an autonomous taxi to pick you up. If you see 10 years later, there will always be a self driving car to pick you up. How can you solve this challenge?
Mavakana: you remind me of when I went to law school. I lived in New York and my mother was in Atlanta. When we talked on the phone, I paid for the long-distance call, which was a lot of money. I’ve been thinking about it recently. There’s no real change in the phone, right? Obviously, now they have become smart phones, but the reality is that the whole business model is about how I can use my smart phone to keep in touch with my relatives without charging monthly fees and long-distance fees, without these restrictions? I think the question you’re asking is, how do you make sure that what you’re creating is pleasant enough to make people feel like a change and an innovation? This is our concern.
It was a consistent experience when I was in waymo. We own this team. I took a Jaguar i-pace and it was a perfect trip. I know what will happen and play my favorite music. If we look forward to the next few years, what will that car tell me when I get on the bus, because I always use it. Does it know my route? Where do I tend to go on Friday night? What do I like in my coffee? From the perspective of commercialization, there are too many things to think about, which really excites me.
Q: so this is an exciting vision, but I want to ask another vision question. In other words, cars in the future may have no steering wheel, screens everywhere, and advertising to you. Is all this to increase Google’s advertising inventory?
Mavakana: This is definitely not something I spend time focusing on. We are working hard, focusing on autonomous driving technology and business model. When I think about the shape of the future car, I think of us attending a code conference in a car, and we communicate across the table in the restaurant. This is exciting for me, so advertising is not something I spend time thinking about.
Q: waymo is a subsidiary of alphabet. What impact does this have on Google?
Mavakana: when waymo was spun off from Google, it happened to be the week I was interviewed. They said, “alphabet will be here and we will support you.” this is my personal experience after joining. Obviously, they are our main investors. We received external funds for the first time last year and introduced more funds again this year. Even if these world-class investors express their confidence in us, it is priceless to have alphabet as our main investor and supporter.
Q: obviously, Sundar Pichai is now the CEO of alphabet and Google. When was the last time Larry Page and Sergey Brin visited waymo?
Mavakana: page and Brin can come whenever they want, but we’re on epidemic prevention, so they don’t come as often as before.
Q: are they very interested in this business?
Mavakana: as always!
Q: I’d like to ask you about supervision. Who are the roadblocks now? Are governments ready for this? Do they have a suitable framework for supervision?
Mavakana: Yes, I think governments at all levels are trying to find a way. I think in the United States, the National Highway Traffic Safety Administration (NHTSA) basically has a voluntary framework. At the state level, each state is somewhat different, but the motivation of most states is to know how to do autonomous driving, how to enable it, how to enable it safely, and have varying degrees of concern. The same is true at the urban and local levels. For example, in Arizona, we have an executive order, and in the first few years, our actions were carried out in accordance with this order. Then at this year’s legislative session, waymo and the policy team took the lead in promoting the legislature to pass the bill and codify it into law. This is only one state, but there are many examples. I think there is a bipartisan consensus that the self driving train is leaving the station.
Q: you mentioned NHTSA. They put forward a crash reporting standard for autopilot cars. Do you think that’s enough? Do you think this needs standardization? If everyone is talking about how safe their car is, does it make any difference to you?
Mavakana: I think it would be ideal to take a comprehensive approach to how the company can establish its own security system. I say this because all our technologies will be different. To some extent, if you set a standard, you say you have to report how you evaluate technical security and make it public. That’s what we do. Last October, we released information about the crash in Phoenix. The reason we do this is that we not only want to build confidence, but also want to prove that our technology performs better than humans under the same circumstances.
Q: both cars and technology are male dominated fields. You are running a leading auto driving company. You are trying to change regulations to make your business successful. How is this for you?
Mavakana: This is the most unforgettable journey of my life. Waymo is really solving problems in a very creative way, which I like. It’s a great team. They want to change the world, so it’s a good way to pass the time. (reviewed by Tencent technology / Jinlu)