Joakim Christoffersson, UX/UI Design Specialist - Volvo

Perfect, thank you. So, hi. I work with HMI development at Volvo Cars. And even though I have a Swedish accent, I work here in Copenhagen in our UX design hub located in Sydhan. And today I will tell you about how we are making a fantastic HMI for the autonomous driving cars of the future. So if you don't know, HMI stands for Human-Mission Interface, or a User Interface, or a UI in short. And for a car, this basically means the interaction between the car and the driver using different types of displays, buttons, stalks, and pedals. And an autonomous driving car is another way of saying a self-driving car. Now, you might be thinking, what do the people who made this know about creating a great user interface? Or a self-driving car for that matter? Well, a lot has happened at Volvo since our boxy designs of the 80s and 90s. And even though I still think this is an awesome looking car, we have come quite a long way since. And actually, our brand new UI might be the most modern and innovative user interface you can find in any car today. We call it Census. Birds fly in the sky, you know how I feel. Sun in the sky, you know how I feel. Breeze drifting on by, you know how I feel. It's a new dawn, it's a new day, it's a new life for me. And I'm feeling good. Fish in the sea, you know how I feel. Blossom on the tree, you know how I feel. It's a new dawn, it's a new day, it's a new life for me. And I'm feeling good. And I'm feeling good. Sleep and peace when day is done, that's what I mean. And this whole world is a new world and a bolder world for me. Sense of the kind, I know how I feel. It's a new dawn, it's a new day, it's a new life for me. It's a new dawn, it's a new day, it's a new life for me. And I'm feeling good. And I'm feeling good. And I'm feeling good. So that was a short video about how we created the new Census UI. We first showed this a couple of months back at the ceremony when we received the award for the most innovative car HMI. The first car we'll find on new Census UI is this fantastic looking beast parked outside. It's called the XC90 and it's based on an entirely new platform we built from the ground up called SPA. SPA means Scalable Product Architecture. This will also be the base for all our next generation cars, including our self-driving ones. And actually the first self-driving cars from Volvo are going to be modified XC90s. But before I start talking about self-driving cars, I'll just take a couple of minutes to paint a picture of the world we're designing for by highlighting a couple of relevant trends. So it probably comes as no surprise to you that people will continue to move into the cities and cities will continue to grow. In 1950, 746 million were living in urban areas. Today, let's close it to 4 billion. In the early 1980s, the city Shenzhen, just outside Hong Kong, was a sleepy little village with a population of 30,000. Today, it has exceeded 12 million and is one of 46 so-called mega cities with a population of over 10 million. Air Quality Index, or AQI, is a number used to measure pollution in the air. This morning's AQI in Copenhagen was 48 and below 50 is considered good air and not hazardous, so you can be safe. But between 200 and 300, that's very unhealthy, and everything above 300 is considered hazardous. During a really smoggy day in January of 2013, Beijing clocked in with an AQI of 755. Of course, that was a really bad day, but Beijing is not the only city which suffers from poor air quality. Every year, 1.25 million people die in traffic accidents, and if nothing is done, that will be close to 2 million by 2020. I know all of you, I guess, have smartphones, and some of you are probably tweeting or checking your Instagram feeds as we speak. And today, consumers are also starting to expect that their cars come equipped with an internet connection and access to the applications they're used to from their smartphones. It's inevitable that we will run out of oil, but it's also necessary from an ecological point that we transition to more sustainable fuel sources. Those were some of the aspects we need to take into consideration when designing a car of tomorrow. So I know Volvo is not the biggest car brand in Denmark, but if there's one thing I think you all know about Volvo is that it's synonymous with safety. This is a quote from Afsar Gabrielsson and Gustav Larsson, the founders of Volvo. And the last part, the part about safety, still holds true today and is a big part of the Volvo culture. The first part, though, the cars are driven by people part. Well, I mean, you can't really blame him for not seeing more than 100 years into the future. So our vision for 2020 is that nobody should be killed or seriously injured in a Volvo car. And our long-term vision is that new Volvos shouldn't crash. Now, I'm going to try a little experiment. I have not tried this before, so I don't know if it's going to work out, but I have a little idea. So can all of you who has a driver's license, please hold up your hands. Now, keep your hands raised if you consider yourself to be at least better than the average driver. So, okay, you can take your hands down now. I should have been seeing half the hands I originally saw. So either we have a severe statistical deviation in here, or it might be that some of you are overestimating your driving skills. But you're not alone in this. It's proven that 80% think they drive better than average. And somewhere here lies our challenge of meeting our visions. More than 90% of all road accidents today are caused by human error. People are speeding, they're driving while intoxicated, they're changing lanes without signaling. In short, crashes happen because of poor driving. So if we can remove the possibility of human error, we can make a safer car. So safety is the main reason why Volvo is investing in self-driving cars. And the other main reason is time. So this is the reality in many cities today. And a few of you might be able to relate to the frustration of sitting still in congested traffic. The average commuter in Shanghai spends almost 50 minutes getting to work. And traffic is then moving along at speeds of about 10 to 20 km an hour. We know that more than 50% of all driving in key Volvo markets takes place at speeds below 40 km an hour. So by making a self-driving car, we can convert what used to be lost time to quality time. Because when you don't have to focus on driving anymore, you can start to prioritize other things which are more meaningful to you. For instance, you might want to send an SMS to someone you care about. Or you might want to read an article about the latest Volvo. But most probably you just do what you always do when you've got some extra time on your hand and you look at kittens on YouTube. So now I've told you about why we are making a self-driving car. Here comes the how. Volvo has now designed a complete system solution. Multiple radars, cameras, a laser and ultrasonic sensors monitor the complete 360 view of the surroundings. A network of computers processes the information, high-performance GPS, and the data. If anything fails, there is a backup solution, just like in an aeroplane. This includes vital components such as computers, sensors, steering and brakes. On the road, the system is designed to be able to control the speed of traffic. The technology is so reliable that the driver can focus on something else without having to pay attention to the traffic. Just like good drivers, potentially critical situations are approached by the driver. The driver can be very careful when driving, and the driver can be very careful when driving. The car reacts faster than most humans. When autonomous driving is no longer available due to exceptional weather situations, technical malfunction, and the driver's ability to control the speed of traffic, the car reacts faster than most humans. When autonomous driving is no longer available due to exceptional weather situations, technical malfunction, or reaching the end of the route, the driver is prompted to take over again. If the driver doesn't take over in time, the car will bring itself to a safe stop. With the Drive Me project, Volvo enters uncharted territory in the field of self-driving cars. Taking the final step to driving safely in real traffic with ordinary people behind the wheel has never been done before. And for drivers, the everyday commute transforms from lost time to quality time, open-form, and safe. From lost time to quality time, open for work or pleasure. The choice is entirely yours. So Volvo's self-driving car initiative is called Drive Me. It is a cooperation between Volvo, the Swedish authorities, and some other actors. And it's really quite a unique initiative. What's so special about Drive Me is that we're taking actual consumers and putting them behind the wheel in real traffic. There will be no trained engineers there to babysit the system. So it's all about the driver and the car. And I tell you, it's something completely different to do this compared to having a concept car drive around the test track by itself. So Drive Me is really the first research project to address autonomous driving from a more holistic point of view. And there are many areas we will investigate. So Volvo's take on an autonomous driving car is to claim back time and reduce accidents. But Drive Me will also look at more societal and economical benefits. In theory, traffic flow can be made a lot more efficient if controlled from the cloud. For instance, if braking, steering, and lane change are controlled, we can optimize flow. However, we have no idea how many cars are actually needed for these to be in a real effect. And the same can be said about infrastructure. Potentially, autonomous driving can change how we design infrastructure. They can be narrower lanes or special roads for autonomous driving cars. A very important aspect is the environmental one. Once the car is aware of its surroundings, it can start to adjust speed and dynamically change routing. According to traffic, topology, curves, and traffic lights, this would optimize time and reduce fuel consumption. A human driver can maybe achieve some of that today, but in practice it's not really done. What's going to be very interesting is to see how the car will be able to drive in a very different way. We will also look at how surrounding traffic interacts with a self-driving car. A potential problem for a self-driving car is other drivers. For example, if a car is not in a good condition, it will be able to drive in a very different way. The problem is that the car is not in a good condition. The problem is that the car is not in a good condition. The car interacts with a self-driving car. A potential problem for a self-driving car is other drivers, because they might not be driving according to laws and regulations. And the final area is the legal aspects. Because, of course, we will not put a self-driving car on the road without having the legal aspects in place. So that's why it's really good that we have the right authorities on board. So on paper, these benefits can be amazing and potentially groundbreaking. But we don't know if they will work in practice. So that's why we have started DriveMe. We want to test with real cars, on real roads, with real users. The first 100 cars, autonomous driving will only be available on certain certified roads. How this will be in the future is difficult to say. These certified roads are roads we have measured in detail, so we know about lane markers, road signs and traffic signs, etc. We will also monitor traffic flow and road work. The first roads to be certified is this 50-kilometer ring road, with connecting roads around the city of Gothenburg. So in case you didn't know, Volvo's headquarters is in Gothenburg. And it's located on the island of Hissingen, up there on the left. So if you follow this road to the left, it will get you to the Volvo facilities. We will also have a cloud solution in place, so we can disable autonomous driving if there is a traffic accident or heavy weather or another event, which would require the user to be in manual control. So now it's time to get a little greasy. Because there is not just manual drive and autonomous drive, and an autonomous driving car can mean a lot of different things. There are five, or what we call levels of automation. And they each have its own unique flavor, and they need to be distinguishable from each other. There are several ways to define these levels of automation. The most commonly used is from the US Authority National Highway Traffic Safety Administration, or as we say in the business, NHTSA. The main purpose of having this definition is to identify who executes the driving and who monitors. So just bear with me for a couple of minutes, because I'm going to get a little bit technical. But it's needed in order for you to understand the area of autonomous driving cars. Level 0. Here the user has full and sole control of the car. This means he controls steering, braking, throttle. Of course, it goes without saying that for level 0, the user is responsible, or the driver is responsible for driving. Level 1 is where most cars are today. Level 1 includes one or more specific control functions, and they work independently of each other. It's mainly oriented around safety. An example is a system Volvo has called City Safety. This system will automatically brake the car before an accident will take place. In level 1, the driver has overall control of the car. And even though some of these systems can assist the driver in normal traffic situations or even crash-imminent situations, the responsibility still lies with the driver. Level 2 includes at least two primary control functions working together to relieve the driver of control of these functions. An example is adaptive cruise control together with lane centering. Adaptive cruise control is like regular cruise control, which you probably know, but it also keeps the distance to the car in front. And lane centering, well, yeah, it keeps the car in the center of the lane. These two together is something we in Volvo call a pilot assist system, or a PA, and can actually be found in the car outside. And this will actually drive the car for you if you're queuing at speeds below 50 kilometers per hour. But the driver is still responsible for monitoring the road, and is expected to be able to take control at a short notice. The main difference between level 1 and level 2 is that in level 2, the driver can be physically disengaged from driving by not having his hands on the steering wheel or a foot on the pedal. So level 3, this is where it starts to get really interesting. Level 3 is what we're talking about in driving a project. And here the driver can hand over full control of the driving through the car, and it's now the car's responsibility to monitor the roads and changes in traffic. The driver is expected to be available for occasional control, but only after a convenient pass-back period. The main difference between level 2 and level 3 is that level 3 is designed so that the driver does not need to make any changes in the road. And it's the car who is responsible. So level 4, that's like the holy grail of autonomous driving. It's mainly about A to B transportation. Here you hand over full control of the car, full control of driving through the car, and the car will drive by itself all the way to your final destination. A prerequisite is that you provide the car with the destination, but after that, the driver will be able to drive to the destination. You provide the car with the destination, but after that, you're not expected to take any part of driving. This also applies when talking about driverless cars. So we're still a very long way from automation in level 4. But we'll, for the foreseeable future, live in a world which combines automation or level 0 to 3. And this is the complicated part. Because users will switch in and out of different levels of automation. And it needs to be clear who's responsible for driving and monitoring. So that was the technical part. So now I'm going to go into something which is probably a bit more familiar to you. And that's the design process. For us, it always starts with the user. Since 2011, we have a strategy called Design Around You, and that's all about people. We need to understand our consumers in order to provide them with a premium experience which is human-oriented while also linking to our heritage of building safe cars. On the left is Kevin. He's our American early adopter. And in the middle is Ye Hong. He represents the new Chinese customer. And on the right, Helena. She's our Swedish comfort seeker. These are not real people. Well, yeah, the people on the photos, they are real. But we have given them fictional biographies. To involve users at an early phase in the design process, we use personas. I'm sure many of you here have seen this. And they're probably as many of you favoring the use of them as they are opposing them. A persona is basically a way to communicate research about people. It's not a real person, but it represents a group of people in the real world. And the purpose is to give you as a designer specific individual characteristics to focus on rather than a group of people or lots of different people. When we have created these personas, we use both quantitative data and in person interviews. And for the autonomous drive project, we have collected a massive amount of data from users all over the world to find out about their behavior and what they expect from a self-driving car. Personas are mainly useful in the early phase of design. And they are very useful in the early stages of design. We also create user journeys. They highlight key interaction points which are relevant to autonomous drive. You can say that the user journey is like an interactive experience. And that's what we call a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. And we have created a persona journey. In this year, we have analyzed how commute situations look in different parts of the world. Because Volvo is a global brand and we need to make sure that our design solutions does not only apply to the Swedish market. What's important when analyzing the commute journey for us is to identify how much time will people spend driving in an autonomous driving mode. And what do the mode transitions look like when switching in and out of this level of automation that I talked about earlier? So, personas, user journeys, hard data. They are all very important for the design process. But what's equally important is that they help us in Volvo to communicate a common understanding about whom we are designing for. And why we make our design decisions. Because in a big corporation, there are as many wills as there are people. And we need to try to put our egos aside and all aim for the same goal. So, now we have a really solid base to start from. We know about the world we are designing for. We know about the brand we represent. We know about the project goals. We know about our users. And we also have some understanding of the technical possibilities and limitations we have to work with. So now, based on this, we can start to formulate some design drivers. And I'm going to let a very short little clip, which is pretty famous, introduce our first design driver. Wow! So, that was Mickey's trailer, 1938. And I think it's pretty impressive that Disney, already eight years ago, could foresee mode confusion. Even though Goofy was driving a Model T Ford. So, as I touched upon earlier, drivers will jump in and out of different levels of automation. And it's important that we know which mode is active. From a driver's perspective, for instance, a level two automation can be experienced as very similar to a level three. In both cases, the car is handling steering, braking, and throttle. But the responsibilities are very different. And the driver might be in for a very big surprise if the controls are handed back to him suddenly without the expected pass back. Because he was thinking that he was in level three. So, designing to avoid mode confusion is one of the most important aspects we need to take into consideration when creating a UI for a self-driving car. To introduce the second design driver, I will use some assistance from Will Smith. What does this action signify? As you entered, when you looked at the other human, what does it mean? It's a sign of trust. It's a human thing. You wouldn't understand. My father tried to teach me human emotions. They are difficult. So, it appears that Detective Spooner and Sonny are having some trust issues. And when in an autonomous driving car, we don't want any trust issues. We need to design a system the user can trust. The trust can be a lot of things, but from a UI point of view, it's mainly about communicating what happens around the car, and why it's behaving the way it is. So, we cannot assume trust from the beginning, but we have to earn it. And we do that by communicating just the right amount of information so the user can feel safe. We also don't want the user to feel the need to continuously have to monitor the system, because after all, the end goal is to free up time for other things. We also have to acknowledge that as the user becomes more familiar with being in an autonomous drive, the need for information might change. Once the system has earned the user's trust, we can start to introduce other types of information or methods to consume content. So, unfortunately, I don't have a good little clip to introduce the third-design driver, but that one is about predictability. From a UI perspective, predictability in an autonomous driving car is about communicating how the car will drive, what will happen, and when. It's about providing reliable information about the autonomous drive road, and especially when autonomous driving is about to come to an end. It can also be about providing information about the autonomous driving system, and it can also be about providing information about which route the car will take, or to prepare the user for what will happen after autonomous driving has ended. So, now I've told you where we're coming from, where we're going, and how we're going to get there. You've been given a look into our design process, and I hope you find it useful and applicable to whatever you might be working with. I'm going to leave you with a serious cliffhanger, because unfortunately I can't show you any detailed designs today. I can tell you, though, that tomorrow I will be at our main proving ground in Hellered in the deep woods outside Gothenburg to perform the first user tests on our brand new autonomous driving UI, and I'm super excited about that. I can also tell you that later this week we will release a new public video, which will give a glimpse into what Volvo has planned for autonomous driving in the future. So, during lunch I will open up the car, so feel free to pop in, have a look, and play around with the Census UI. Thank you. See you on the road.