Bicycle Commuting

MOTIVATION

Statistics show bicycling has been growing as a mode of transportation in the last decade. Specifically in the city of Atlanta, bicycling to work grew 180% in the last 10 years – bicycling now making up 0.8% of all commutes in the city. However, looking at the demographic breakdown across the nation, men outnumber on average women about 3 to 1 in bicycle commutes.

Prior research has also shown a gender misalignment in the decision making of whether to commute by bicycle or not. An article written by Welke et al. in 2004 explored women’s cycling barriers in several countries such as Canada, Mozambique, the Netherlands, and China. The article stated that, within the City of Ottawa, 50% fewer women cycled to work or school than men (0.7% vs. 1.6%). In Continental Europe, women make up between 10-23% of bicycle commuters. A significant amount of Ottawan women (65%) were more likely than men to agree that that traffic safety was a major barrier to choosing to cycle. Understanding why women are making less trips than men can help gain insight on how to increase the number of trips made to work, lowering barriers of entry for women riders, and or improving the ride experience.

SURVEYING CONFIDENCE

A 13-question online survey was deployed to better understand bicycling habits and needs of the working population in Atlanta. The survey focused on different modes of transportation, distance from home to work, workplace accommodations for bicyclists, and people’s concerns with cycling to work.

• The survey garnered over 160 responses (44% female and 56% male). The average respondent’s age was 27 years of age with a standard deviation of 9 years

• While 46% of survey takers felt that biking in general is unsafe, 70% felt that biking in Atlanta is unsafe

• Respondents were asked to rate themselves and their experience as a cyclist on a Lickert scale. (1: “I am unable to bike,” 5: “I’m an expert”)

• Both males that biked to work and males that traveled with another mode of transportation tended to rate themselves higher in experience than women did.

The survey contained a free-response section in which respondents were asked to list all of their concerns with bicycling to work. Using affinity mapping and post-its to list all concerns, there was noticeable patterns of women mentioning safety in their response.

After clustering and regrouping the post-its, I noticed that the male pool shrank significantly more than the female pool. This led to a possible insight of: despite personal concerns issues, most men cycled regardless. In addition, I was able to gain insight on the issues women specifically had with bicycling to work. When I pulled all responses that used the word “safety”— to observe if there were additional safety concerns amongst women— nearly every issue fit into one of four categories: Confidence, Driver Awareness, Space, and Visibility.

ISSUES

INTERVIEWS

Eight women were interviewed and followed up with throughout the duration of the project. The 8 women’s ages range from 22- 50 years of age. They all worked inside the Atlanta perimeter. The sampling of women included both frequent and infrequent cyclists. The frequent cyclists would ride to work as often as 5 days a week. Infrequent cyclists would ride as less as once a month.

The interviews of female riders were used to further delve into what steps or thought processes were involved when deciding whether to ride. The interview format was comprised of a series of semi-structural questions for the subjects to answer. The interviews were audio taped for documentation and transcription.

What I found from the women was similar to the findings of my survey. The women wanted space (3-6 ft of space between them and motor vehicles) and for cars to see them. For the infrequent cyclists, most wanted more confidence in themselves to ride with cars.

DESIGN OBJECTIVES

The following design objectives were developed to help design a product to get women back on the road:

1. Increase feelings of safety while commuting by bicycle

2. Increase driver awareness of the cyclist to create more physical space

3. Increase perceived space to create rider confidence

DESIGN DEVELOPMENT

In addition, they didn’t want any type of physical intrusion on the road. And several agreed that while light-up apparel seemed pretty novel, they themselves would not “get caught dead” wearing it riding to work. As one girl put it, “design me an invisible force shield so that cars will go around me.” So I went back to the drawing board.

In order to utilize vehicle-to-vehicle technology, the second series of concept development comprised of concepts that were 1) large enough to house V2V technology and 2) something more than just the V2V technology. In order to not reduce additional space on the bicycle, the product double as something such as a bicycle light or bicycle computer.

DESIGN DEVELOPMENT: Transceiver

Next, I developed a series of study models, varying in shapes and sizes. I used playing cards and batteries to simulate the area needed for V2V (the technology being about the size of a standard cigarette pack). While many of the women tended to favor the smallest model, it was unable to contain the technology, power source, and circuit board for LED lighting. These study models helped determine and better guage the height, width, and depth constraints for the product.

The study models were then taken and attached to a bicycle to see what the device would look like as a back light on a bicycle. This also helped better grasp proportions and front profile shapes to better complement the bicycle.

USER GENERATED MODELS: Retrofit Receiver

Using styrofoam, I created a collection of abstract shapes and asked subjects to build a receiver that would be located in a car. As instructed to the subjects, the receiver’s purpose is to show the driver the location of nearby cyclists. When given the pieces, 4 out of 8 of the women picked up the square piece and insisted that the device would be flat and easily mountable on the windshield. The explanation for the top and bottom images to the right were developed by two different women that said that their device was inspired by a compass. The top model would show the bicyclist being north, south, east or west.

FORM REFINEMENT

I then took the study models and developed a generic profile for each orthographic view (top, side, front). Over 50 iterations of each orthographic for both the transceiver and retrofit receiver were drawn. Following the iteration process, I presented the iterations to several people and selected the ones that tended to draw people in more. Using those, I geometrically refined each of the orthographic views. Seen below is one version, the front view, in which each unit is based off of the width of the LED casing.

INTERFACE DEVELOPMENT

The purpose for developing an interface was to find a way to quickly convey information from the bicyclist to the driver. Because driving requires attention to the road, it’s important to inform the driver concisely. We took a few cues for inspiration by looking at threat condition maps (theatcons) located in the bottom left corner of several first-person shooter games.

Threatcons give the game player relevant information on moving targets within the player’s environment. Because the game player is focused on stayed alive and not getting hit, the threatcon’s purpose to to convey all the necessary information on nearby targets in a glance. Similarly, because cars and bicycles are both moving targets, the environment and subjects in a specific vicinity are changing at any given point in time. Because of this, the interface needs to be able to convey the necessary information about nearby cyclists in a glance. After all, the driver’s eyes should first and foremost be on the road.

I then created and printed several different basic interfaces to be used in locating bicyclists. The tick marks would indicate a cyclist’s position in relation to the driver (the center of the graphic). When I presented these to the women and other male subjects, the ones that tended to be favored were ones that could give a more accurate location of a cyclist. The graphics with more tick marks would also be more helpful for drivers in a real-time scenario of tracking nearby cyclists.

From one of the interfaces that received positive feedback , a quicktime mockup video of how it would relay messages was generated. This was important as the previous mockup was static and subjects could not see how it would work. When the video was shown to the subjects, we realized several things. First, it was unclear to whether the tickmarks were referring to the cyclist or the driver as the center symbol was a bike. The center was supposed to represent the driver, but because the symbol was a bike, subjects were unclear to whether the moving tick mark was them... or the driver. Second, this model could not convey how close the cyclist was to the car. Was it 60 feet within proximity? Or 7 feet? And was there only one cyclist present? Or three?

Using the subjects’ suggestions and the findings from the last followup, I refined the interface to consist of three concentric circles of points around the car. The concentric circles represent zones within a vicinity of the car. The outer zones being less imperative for attention while the closer zones showing the driver that the bicyclist was very near. This design also allowed us to convey whether several or one bicyclist was in the vicinity of the vehicle. And it also conveyed more information about the cyclist’s location in reference to the driver in a quick glance.

Because a vehicle passing a cyclist tends to be more life-threatening to the cyclist than when the vehicle is in any other position, the interface zones were refined to illustrate priority and urgency. The top image to the right highlights the areas a cyclist may be present when a vehicle is passing the cyclist within a distance of one lane. The outer circles to the side tend to be less urgent as cars two lanes over tend to pose less danger to cars riding in the same lane as the cyclist.

DETERMINING DEVICE VICINITY

When interviewed, most women gave a range of 20-40 ft for the desired range in which they would like to appear within the tracking vicinity. Stopping distance = thinking distance + braking distance The Stopping Distance Chart was used to help determine device vicinity and safety zones within the receiver’s range. According to the chart below, if a car is closing in on a cyclist at 20 miles an hour, the driver’s thinking distance is 44 ft. If the device’s vicinity is a 40 feet, by the time the driver reacts, he or she will have theoretically passed the cyclist already. Because of this, 40 ft of range is not sufficient enough. To leave enough time and distance for the driver to think, react, and adapt, we are proposing a range of 150 feet for the receiver to pick up cyclists’ locations.

A PASSIVE DEVICE

In determining the what information should be exchanged between the bicyclist and driver, I asked the bicyclists if they wanted a passive or an active device.

A passive device is one where the user and turn on the device and no commands are necessary for incoming or outgoing messages. An active device would be one where the user would key in commands or prompt for information or updates. When asked whether they wanted to receive information or feedback from the vehicles, most of the women were against it.

DISCUSSION

This study only covers a conceptual level of the Bicycle Awareness System. The results of the project were based off of the subjects’ perceptions and understandings of the technology and the system. Actual testing and being able to trial the device through traffic would help the system develop further.

In determining the vicinity of the device, there is room for improvement. Without testing, it is difficult to determine how drivers might react to the receiver’s alert. With testing, we may find that 150 ft may not be the optimal amount of distance to alert the driver. There may also be the possibility of needing a greater range of tracking on roads with higher speeds.

There is also the issue of the size of the transceiver. Currently at it’s size (6 inches in width and 4.5 inches in depth), the transceiver is large and heavy. If the technology can improve and the size and weight of the transceiver could cut down, the product itself would be more marketable.

Source: Georgia Tech Archives (2012)