Samsung DRVLNE

arrow pointing up
Back To Top

Overview

Context

Samsung is a leader in consumer electronics but has yet to enter the automobile space. A huge market opportunity, autonomous driving is the future yet Samsung wasn’t sure how it could provide value. Being a leader in software and screen technology, it only made sense that Samsung look at vehicle dashboards as a place to innovate and impact autonomous driving.

Samsung at the time developed a semi-conductor technology to run autonomous driving systems. This platform needed a UX/UI for the users to control and understand. This system needed to be adaptable to work with technology today (Level 3) as well as scale to the future (Level 5)

Outcome

This project outcome was a video to demo the use cases of the Samsung DRVLNE dashboard and high fidelity mockups to show dashboard capabilities which was showcased at the Harman booth at CES 2018.

Deliverables

  • A UX Spec which would outline the design principles, interactions and specify expected behaviors for this system
  • A 3-5 minute promotional video which would show how this system would work for different common day to day use cases like collision prevention
Role
Lead UX Designer
Timeline
3 months
Team
Myself and 1 Designer
Company
Samsung

At the time of this design autonomous vehicles were considered a Level 3 in autonomy. Because of that there were very specific considerations and challenges when design in:

Situational Awareness

Taking control of the vehicle requires you to have some understanding of what is going on at that mo-ment. If a driver is distracted and not paying atten-tion to the road, when they take over they may make more mistakes then if they were paying attention. “The issues underlying difficulty in taking control of a vehicle are allocation of cognitive resources and situation awareness” (Young et al. [10]).

Cognitive Resources

Humans have limited cognitive resources. If they are reading something, their visual resources are limited, and thus a visual alert, will be ineffective. Alerts and communication from the car must take into account the drivers cognitive resources in use at the time and decide alternatives to ease the strain of balancing multiple cues.

Over-Reliance

The average person has different assumptions about what autonomous driving means and what the limita-tions are. It’s critical that the design system lets the user know what they can and cannot do in the vehicle. Drivers should not over or under trust the system.

Trust & Perception

In general, users still distrust autonomous driving systems which stifles adoption.

The bigger questions we want to solve

question mark

How might we create a driving system where the driver can be in different stages of attention or awareness?

question mark

How might we ease fears while not overpromising on what we can deliver?

question mark

How might we create a driving system that is adaptable to the environment and the unexpected obstacles or opportunities?

Research and Workshoping

Being that this was a new space for Samsung and for myself, I dove into the tons of existing research from academic sources to learn more about autonomous driving systems and HCI.

I found that there is an incredible autonomous driving research hub at Standford University and organized an on-site visit to speak with professor and leading academic in autonomous driving Wendy Ju. Here we were able to discuss some of the common challenges and learn ways to mitigate them using UX design.

From the research I was able to synthesize the findings into frameworks and diagrams

Mental Modals

Below is a mental model map of how we imagined users using an autonomous versus manual driving system. Thinking about the users body, posture and mindset we laid out the way that they would use and think about the dashboard system.
mental model chart

User Modes

We noticed that users are very different depending on if they are in the driver’s seat or if they are in an autonomous vehicle. We wanted the system we create to either be an Enhanced Drive: meaning if you do have to drive it’s easier, pleasurable and still leverages sensor technology to make you a better driver. Or if you are in autonomous drive, we called it Supervised Autonomy. In supervised autonomy the user changes roles into more of a supervisor who only has to occasionally monitor the driving system in case of emergency.

Car Persona

We wanted to think about a simple way to describe the value and skill level of the system we were creating. It is not level 5 autonomy where it needs no intervention, so we imagined this would be similar to a Co-Pilot in aviation. With a co-pilot a pilot can take breaks and rely on them to takeover control in certain situations. But the role is more of a peer versus a lead.

car persona

Predicted Overrides needed:

We also anticipated there may be times a user attempts to override the system in various ways. We wanted to preempt that and plan for those instances so that it doesn’t cause an accident. These are the types of overrides we anticipated:
  1. Grabbing the steering wheel
  2. Stepping on the brake intending to stop
  3. Changing navigation
  4. Merging into a different lane

Opportunities

After synthesizing the research and thinking about the user needs in a semi-autonomous vehicle I identified 3 opportunities for design and product to differentiate us from the competition.

  • Autonomous driving capabilities are a spectrum. There are many factors that have to be aligned in order to safely drive autonomously now or even in the future. We need to be realistic and let users know when they can use it to ensure safe adoption and success using new technology.
  • Plan for structured and unstructured handoffs between the user and the autonomous driving system.
  • Even in autonomous vehicles, there needs to be ways to allow the user to intervene to help build trust and comfort that they have control if the choose to.

Jobs to Be Done

As a driver, I want to know when I can use autonomous driving system safety, so that I can do other things while driving or take a break.

As a driver, I want to be able to drive safely, so that I can have a more enjoyable experience driving to my destination.

As a driver, I want to be able to know how to intervene in the autonomous driving system, so that I can stop it if I want to take-over.

Architecture

Information Architecture

Below is a mental model map of how we imagined users using an autonomous versus manual driving system. Thinking about the users body, posture and mindset we laid out the way that they would use and think about the dashboard system.
Information Architecture

User Flows

Autonomous Handoff UX Flow

With the design I created user flows to outline the expected behavior for a smooth handoff from autonomous to enhanced drive.
Autonomous Handoff UX Flow
The flows would take into account voice and visual UX to alert the driver of what is happening.
voice and visual UX to alertvoice and visual UX to alert image 1

Design + Key Features

Alerts

Alerts surrounding autonomy and autonomous drive are critical to a safe Level 3 experience. Below are some examples

Alert

Situational Awareness Alerts

Alert

Autonomous Alerts

Alert

Enhanced Drive Alerts

Notification timing

Notification timing

Iconography (Key)

iconography

HUD

A heads up display to help the driver know critical information as they are in Enhanced Drive. Below gives a glimpse into the design components and features for the head’s up display
Autonomous
Autonomous HUD
Manual
Manual HUD

Dashboard

The dashboard in manual mode should look like what a dashboard traditionally looks like to drivers, so that if/when drivers need to drive, there is no learning curve to learning how to read the dashboard. 

The dashboard in autonomous mode helps the driver supervise their car driving. As we get closer to fully autonomous driving, the steering wheel may disappear and the need for a dashboard may dissipate, but in the meantime a dashboard that can dynamically support the driver in manual and autonomous will be critical on the path to full autonomy.
Autonomous & Manual dashboard

Autonomous Indicator

An innovative feature we designed was an Autonomous Indicator. This feature ended up winning a patent. The autonomous indicator is always on and lets the user know if autonomous drive is available and safe to use. It uses sensor and map data to determine how much human supervision is needed for the drive. This is important because setting expectations with the user allows them to make the best decision for how to interact with their vehicle.  

The autonomous indicator will let the user know in time how long they have before the system can no longer drive for them. If a user does not take control by this time, the vehicle will safely pull over and park.
Autonomous indicator Key
Autonomous indicator as displayed on Dashboard
Autonomous Indicator & dashboard

Dash+

Where the dashboard provides drivers car diagnostic data, the Dash + can serve as a more robust view of the driving environment as well as an input method for the autonomous drive. Dash+ UI should be visually different in autonomous and manual mode. But the dashboard and Dash+ should use the same design style and language. Dash+ should include safety and infotainment so that in the event the driver is watching content, safety alerts can overtake any infotainment that is taking place. 

A center console screen that can be used for autonomous control and/or entertainment when in autonomous drive mode. 

This surface becomes important when the driver is in a leaned back and relaxed state and may not see the dashboard or HUD for important info.
dash plus manual and autonomous
Autonomous Indicator & dashboard

Dash+ Prototypes

In this video we demo how the start up handoff interaction would go. Here the driver tells the system where they want to drive to and detects that this trip can be driven autonomously. You see how we use sound, visual to alert the driver when autonomous is driving or when the driver needs to take over. 

This is another example of the system telling the user it needs to take over. This video shows the mechanism of pushing buttons on the steering wheel to confirm driver is ready to take over driving and accepts the handoff. 

Results

The final result was we presented a video to demo the autonomous driving solution and shared it at the Samsung/Harman Innovation booth at CES 2016.We got alot of positive feedback and solidified our partnership with Harman on autonomous driving. Samsung now has a point of view on autonomous driving, a huge win from when we started at the beginning of the project

View My Work

group of managers and workers

Sales Enablement

This is a block of text. Double-click this text to edit it.
View Project
sales enablement

Einstein Conversation Insights

This is a block of text. Double-click this text to edit it.
View Project