Rectangle-26

THE ULTIMATE SAFETY BELT

 Saving San Diego's "Fatal Fifteen"

Rectangle-26

PROJECT TYPE

Civic Design

MY ROLE

UX Researcher

Product Designer

TOOL

Adobe Photoshop

Adobe Illustrator

iMovie

Cinema4D


TEAM

Dean Ravida

Erich McMillan

Joel Yun

Lindy Wong


THE ULTIMATE SAFETY BELT

A smart seatbelt that alerts unexpected pedestrian movements, in cooperation with the smart streetlight technology.

THE CONCERN

200+ were killed or injured at Fatal15.

San Diego has named 15 of the most dangerous intersections based on the number of traffic accidents. More than 200 people have been killed or seriously injured at these fifteen intersections over the last 15 years. The majority of the accidents that occur in these intersections are between pedestrians and vehicles.

FATAL15

Overlapping satellite photos of San Diego's "Fatal 15" intersections

FIELD OBSERVATION

Pedestrians aren't visible.

Observed:

  • Pedestrians who jaywalk tend to check for cars more often than those who cross according to traffic signals.
  • Pedestrians may walk in front of cars through crosswalks without checking traffic.

 Different things obstruct the driver’s vision of pedestrians:

  • Large vehicles ahead of you to the left and right lanes block the view of pedestrians at the intersection.
  • Parked cars on the sidewalk (horizontal and diagonal parking).
  • Ongoing construction.
  • High traffic at peak hours.

Footages at San Diego's "Fatal 15"

PRELIMINARY IDEA

Between-Streetlight Pedestrian Signals

  • LED bar over the street.
  • Passive Infrared motion detectors.
  • LED Pedestrian crossing signs.
  • Function for real-time pedestrian tracking and displaying over the street.

3D demonstration of the "Between-Streetlight Pedestrian Signals"

The above idea is also our entry in the "D4SD: Design Forward 2017" design contest.

PROBLEM W/ THIS IDEA

But...

  • Low visibility conditions make LEDs invisible.
  • Possible distraction or confusion with other lights on the streetlight.
  • Installing the system would likely be expensive and intersection-specific (due to varying road widths).
EJXFUUba1150iNaVzfQX2hlct4j9-WGw6Q2On2dSVp2sPAaAVmscN_Tkz_2L2n2MQ7YFFyQ905zvcUzoqEEZbJ3nUlrlQYuia6wz__m7Ei9zGCvuQFXsJQOx7FWqi04vBcME4-4s-No

Low visibility caused by bad weather

NEW RESEARCH DIRECTION

What about...
Into the car?

  • Notifications inside the car are unaffected by weather conditions.
  • Smaller devices to be implemented into cars will be cheaper and easier to install.
  • With smart streetlight technology to be implemented in San Diego, real-time pedestrian tracking data will be available.
COPYRIGHT CHRISTOPHER COOPER copy
Research project Augmented Reality – contact-analogue Head-Up Display (10/2011)

Heads-up displays for cueing??

Not really.

  • Visual signals within the car require drivers to shift visual attention away from the road and into the car, acting as a distraction.
  • After shifting attention, drivers experience attentional blink when they are less attentive to the rest of the road.
  • Augmented reality HUDs are only as effective as they superimpose their graphics onto the road effectively, visually appearing as part of the world. 
    However, weather conditions like fog interfere with the illusion augmented reality technologies rely on.

POSSIBILITIES EXPLORATION

How do we cue the drivers without distracting them?

The scientific literature says...

  •  “Tactile warnings are more effective than auditory warnings during both simple and complex conversations [in the car].” (Mohebbi et al., 2009)
  • “Drivers detect fewer signals in the central visual field when driving slowly and it appears that they detect fewer signals in peripheral vision when driving quickly.” (Rogé et al., 2004)
  • Vibrotactile cues are effective at cueing visual attention, especially when correlated to the direction such as right versus left cues. (Ho et al., 2005)

 

Blank Diagram-3

Analysis of different stimuli types

NEW IDEA

Seat Motion Alert

  • Add lower back vibration pads to the driver’s seat.
  • Pedestrian motion data collected from Smart Streetlights are sent to the receiver in the vibration pads.
  • Motion pads either vibrate on the left or right side depending on which side the pedestrian is coming from.
  • Users can subscribe to data via an app:
    1. The app may also be used to modify vibration pads (e.g. turn them on/off, adjust the vibration intensity, etc.).
    2. The app may also be used to help pinpoint the user’s location and connects the prototype to the local pedestrian traffic).
Processed with VSCO with v6 preset

Processed with VSCO with v6 preset

Smart streetlight technology

Smart vibration pads on carseat

POSSIBLE PROBLEMS

Still not perfect?

  • It takes time for drivers to associate the vibration with the direction of pedestrians coming - stimuli not intuitive enough and learning curve exists.
  • Different seating positions/body shapes might cause inaccurate stimuli.
  • Effects easily influenced by texture/thickness of the clothes.
  • Too much vibration might cause drivers to subconsciously not lean back on the seat.

FINAL SOLUTION

The Ultimate Safety Belt!

  • Pedestrian motion data collected from Smart Streetlights are sent to the receiver in the seatbelt.
  • The seat belt either fastens from the left or right side depending on which side the pedestrian is coming from.
  • Users can subscribe to data via an app:
    1. The app may also be used to modify seatbelt (e.g. turn them on/off, adjust the fastening intensity, etc.).
    2. The app may also be used to help pinpoint the user’s location and connects the prototype to the local pedestrian traffic).
Processed with VSCO with v6 preset

Processed with VSCO with v6 preset

Smart streetlight technology

Smart alerting seatbelt

PROTOTYPE TESTING

Test it out!

For the prototype testing, we:

  • Chose the parking lot as our testing area, for its high density of "blind spots" (and lower speed as well because we want to keep our participants safe!);
  • Chose the evening as our testing time, for its low visibility;
  • Had person A hold a detached seatbelt and sit behind the driver so that person A can simulate how the device works;
  • Had a "pedestrian" person B run out from one blind spot on the left of the driver's right before the car drives by, and let person A pull the seatbelt from the left to create a stimulus for the driver (Note: person A knows where person B would run out).

Prototype testing

Thanks for reading!


CONTACT

© Siyuan Gao 2020