Waymo
Pedestrian Communication
Pedestrain Communication
Waymo began as Google's self-driving car project in 2009. Their goal is to make it safe and easy for passangers to get around but have they considered everyone on the road?
I worked as part of a multidisciplinary team that consisted of a copywriter, art director and experience designer. While I was at VCU Brandcenter, we were asked to build trust with an emerging technology so we focused on autonomous vehicles during the transition period before mass adoption.
Our final deliverables were a campaign and a minimum viable product that helps pedestrians feel recognized and understood. I was responsible for market research, prototyping, user testing and visual storytelling.
My Role
Research, Prototyping, User Testing, Interaction Design, 3D Design
Team
Zach Vono, Mitchel Moss, Collin Oshea
When pedestrians encounter vehicles, they expect a signal from the driver to see if they should walk or wait.
If the driver sees them and wants to let them walk, they let them know by smiling, making eye contact or giving a wave. When drivers are removed from the vehicle, it's the vehicle's responsibility to communicate with pedestrians.
Right now, there are hundreds of Waymo vehicles driving around California and Arizona without a means to communicate with pedestrians. 'Waymo One' is the pilot program for the company's commercial ride-hailing service. Some of the most popular trips taken by early riders are places are to work, school, and the supermarket. These are all places that are crowded with pedestrians.
Unmarked crosswalks, suburbs, and rural environments present higher risks for pedestrian safety.
Pedestrian Communication System
The communication system lets pedestrians know when they are seen and when it's safe to cross the road. It uses familiar symbols to avoid language barriers.
Product Demonstration (No Audio)
How it works
The ICS only activates when a pedestrian intends to cross the road or a biker is on either side of the vehicle. It adds transparency to how Waymo's vehicles think and behave, making them more predictable so pedestrians know what to expect.
Features
1) 'Walk' and 'Do Not Walk' Symbols for Pedestrians
2) Light Bars for Bikers
Design Guidelines
Glanceable
- Communicate the information clearly and without distractions
- Always display the vehicle's current information
- Place signals near the front windshield to maximize scan-ability
Actionable
- Anticipate the user’s needs by displaying timely and relevant information
- Use background time to update fetch new data and refresh the display
- Clarity, not clever
Responsive
- Interactions with the self-driving car should be quick
- Respond to pedestrian interactions with immediate feedback
- Minimize the time it takes to launch and load new signals
Research
Pedestrian decision making
After spending time at intersections and speaking with pedestrians, I wanted to dive deeper. My goal was to understand the moment of meaning between perception and action. It felt crucial since pedestrian communication impacts both the safety of the public and the perception of Waymo's brand. My next step was researching the factors in pedestrian decision making to understand the different variables at play when pedestrians encounter Waymo vehicles.
Empathy Map
After user interviews and researching factors in pedestrian decision making, I constructed an empathy map to create a shared understanding of what pedestrians experience at crosswalks and intersections. I was able to identify the user needs based off of the most frequent patterns within the data.
Prototyping
Flowchart
Next, I created a flow diagram that adds pedestrian communication to the vehicle's current functionality. Pedestrian communication consists of outward messages about what the vehicle sees and intends to do.
Prototyping
Research Goal
Determine the most effective way for driverless cars to communicate with pedestrians.
Hypothesis
If the ICS uses a combination of symbols and colors, then pedestrians will know when it is safe to cross the road
Approach
I prototyped two versions of the ICS in After Effects and tested them in a survey by animating an existing image of a Waymo vehicle at an intersection. The first set used only color while the second set used color and symbols.
First Prototype
Second Prototype
Findings
The results varied but it was clear that a form of context was missing for the first prototype. A little over 50% of the respondents thought red meant they were safe to walk and green meant the vehicle was about to go.
What worked?
The colors from the traffic signal offer a form of direction and placing the system near the driver windshield indicated that it was for pedestrians.
Moving Forward
Add the pedestrian walk signal (without the orange warning color) to the equation.
Key Takeaways
Over 90% of the respondents understood what the vehicle was trying to communicate. Self-driving cars should communicate with pedestrians by using a combination of symbols and established traffic colors.
