Join us to participate in the upcoming 2019 City:One Challenge. 

Crowdsourced Accident Risk Input For Spotting Higher Risk Vehicle Collision Locations

A patented app which monitors drivers for harsh braking, aggregates locations of harsh braking, and displays clusters of them on a map

Photo of John Lindsay
4 1

Written by

Pittsburgh has ranked among the highest for frequency of crashes. The question is how to decrease the number of crashes. Hard braking is considered one of the most highly predictive variables for predicting future crashes. Normally harsh braking is tied to individual drivers. This application ties harsh braking events to location so that clusters/"hotspots" may be observed at certain locations over time.

This composite app has two parts. The front end app is a smartphone app which monitors drivers for harsh braking. When harsh braking occurs, that location is logged. Over time clusters of harsh braking may be observed on the map display, enabling individual drivers to change behavior at those locations. The back end is an anonymous, aggregate view of harsh braking events across drivers who choose to contribute to the city.  A searchable map display of those aggregate events is available to the city traffic engineer. Again, clusters of harsh braking may be observed for the attention, with higher frequency clusters drawing attention of the traffic engineer for possible corrective action.

Describe who will use your solution (1,000 characters)

Individual drivers, such as citizens or city fleet vehicle drivers, would use the front end application. Traffic engineers or fleet managers would use the aggregate, back-end application.

Describe your solution's stage of development

  • Prototype - you have built a prototype and tested it with potential users

Tell us about your team or organization (500 characters)

John Lindsay - application developer, inventor ( https://stopclusters.com/ )

Size of your team or organization

  • I am submitting as an individual

Funding Request

  • $50,000

Describe how you would pilot your idea (1000 characters)

The expected use of funding would be to fund further develop of the mobile and backend applications, the server storage and bandwidth, and promoting/incentivizing the app to drivers. The test would be whether users 1. will use the app and 2. do so within reasonable incentive cost. After a pilot period of feedback, usage, and feature revisions/additions, the expected fee structure would monthly or annual fee. The anticipated fee structure is that the front-end app would be free to drivers. The municipality or fleet owner using the back-end would be the payees. The hard money ROI for a municipality would be a reduced frequency and man-hours of EMS deployments. Soft money return on investment for a municipality would be factors such as increased safety reputation by not being known as a high crash frequency city. Marginal EMS deployment costs vary widely, with severity of collision being a key factor but the goal is to charge less than the cost of the three EMS deployments.

Describe how you would measure the success of your pilot (1000 characters)

Key metrics are expect to be: Number of front-end app installations Frequency of front-end app use Number of harsh braking events measured Frequency of back-end usage by traffic engineers Optimally, number of hotspots targeted for correction Optimally, percentage accident reduction upon correction

4 comments

Join the conversation:

Comment
Photo of Idrees Mutahr
Team

Hi John, I am a facilitator for this project, I think this is an interesting idea! What I think would really solidify the value of this proposal is if you could articulate the resulting uses of the heat map more clearly, how cities will use the maps and what they can change about those harsh braking locations to reduce crashes.

Photo of John Lindsay
Team

Hello Idrees,

This app is not intended to supplant the knowledge and options by traffic engineers/traffic study, rather it is primarily intended to help draw their attention to the "hotspot" in the first place (ie somewhat of a predictive analytic). Then, it can aid in their root cause analysis to the potential problem.

To illustrate, I've added a zoomed out view with a data snapshot from Buenos Aires to the media section. First, a "traffic engineer" would periodically view the aggregate data for a time period to see if there is anything that warrants additional attention. The single map markers would not likely warrant attention. However, the 11 count cluster stands out, as it is a multiple higher than the other clusters for the time period. Thus it may warrant extra attention.

The app can aid in cause analysis. For example, using the time search filters on the left, they may see that all 11 harsh braking events occurred in a single night. This may indicate, or at least narrow, a set of causes. With supplemental searching and/or their knowledge of the city, they may further narrow the causes. For example, as this is a city center, perhaps there is road passing by a concert venue with events at that time. All of this would feed into the traffic study for potential problem and solutions, perhaps leading to something like electronic signs during the events that display something like "event traffic ahead, please slow down." To view the effect of the post impact solution effect, they may monitor that same location post-implementation to evaluate impact.

If the harsh braking events are consistently "one-off" harsh braking events spread over time, that might support a starting inference of something more infrastructural such as upstream traffic flow/speed/timing. If the harsh braking events are consistently "small cluster" harsh braking events over time, that might support starting inferences of something along the lines of pedestrians crossing during rush hour/happy hour.

And sometimes it might only serve to promote some physical/camera observation/inspection of the hotspot cluster. For example, in Dallas there was a problem spot for more than five years where there was a merge right sign but the lane indicators were painted to show left merger. I'd see accidents/near-misses/aggressive driving/swerving/other dangerous driving events often (often involving harsh braking). Simple inspection indicates the source of that problem.

Photo of Idrees Mutahr
Team

That makes sense John thank you for the response!

View all comments