16 June 2016

Open Seattle June 2016 Meeting: Public Safety Data & Vision

Seth Vincent hosted the meeting.

City of Seattle Open Data Policy update

Candace Faber gave the update. She just finished the City’s first data camp. This is the first major milestone of the new Open Data policy – the city’s data is open by preference. She was teaching city staff champions about open data, why it matters, and how Open Seattle members might want to use it.

For more information: Open.data@seattle.gov, City of Seattle’s Open Data Resources

Public Safety data, Robin Jones, Socrata

Robin Jones is a Vice President at Socrata, and leads growth initiatives at the company. Most recently, she has focused on public safety agencies. Law enforcement agencies are facing extra scrutiny recently due to high-profile officer involved shootings and use of force incidents. The White House convened a task force on 21st century policing last year. Two recommendations came from that: Police agencies need to be more transparent, and they need to do better citizen engagement. This lead to The White House Police Data Initiative to promote open data from law enforcement.

Socrata is currently supporting an older app called Crime Reports. This is a crime incident map, and about 1,100 police agencies upload data there. The technology had not been updated since 2004, until Socrata started working on it. Now, police agencies can put in messages related to specific incidents on the map. Users can also access an open data portal . The data is all in the same schema, allowing you to compare apples to apples. There is lots of public safety data for Seattle.

There is also a feature that provides some context that puts data into perspective. Once example is an opioid addiction data story. The app still provides access to the raw data so others can review the data under their own power.

Data and Vision Zero at SDOT (Transportation Safety)

Jim Curtain from SDOT presented on Vision Zero, which is supported by SDOT and SPD. The Vision Zero goal is no transportation-related deaths or serious injuries by 2030. They are making data-driven interventions: making streets easier to cross, reducing speed limits, and many other aspects.

2015 was the first year in a long time that nationally there was an increase (14%) in transportation-related deaths. In Seattle, there was a 13% increase, to 21 total deaths. The trend has been mostly positive over the last decade, however. Vulnerable users like bicyclists, pedestrians, and motorcyclists are over-represented, at 62% of total fatalities. They also have disproportionately high serious injury rates.

Seattle’s Vision Zero uses the data to guide where to invest in interventions. They just got $930 million over 10 years in a citizen-approved tax levy. The data helps them evaluate and improve our streets, and choose the best countermeasures. These evaluations also build public trust.

In the future, Vision Zero wants to be much more proactive. They are currently doing a retrospective analysis on 2007-2014 data, trying to better understand the risk factors, then identify locations that include those risk factors and proactively address them. For instance, early findings show that 62% of serious bike collisions and 70% of serious pedestrian collisions occur at intersections. DataKind and Microsoft are helping continuing the analysis.

Vision Zero has pavement data, collision data, speed data, street network data, topography, transit data, bicycle data, and many other data sets. However, they don’t have good exposure data – car, pedestrian, and bike volumes.

They are interested in an Open Seattle partnership to help with: behavior change, enforcement strategies, engineering, and public health interventions, among others.

Jim took several questions from the group:

Does your data have directionality? The data on data.seattle.gov is a flattened dataset, and doesn’t have everything. Vision Zero does have directionality for most crashes and could provide that data now if requested.

When will the Duck crash (Aurora Bridge) data be released? We don’t know. For major events like that, it takes a long time to get investigate, and results can be limited for a year or longer.

What demographics do you collect in the person-records? Age and gender. We only have ethnicity for fatality data. We have explored a partnership with the Department of Licensing and with medical sources like hospitals, but have not been able to collect it yet.

There are limitations around your street engineering dataset. How far back in time, and what pieces of information can we get today? Such as number of lanes, how wide they are, when changes were made, etc? Processes have been tightened up in recent years, and we just got that data into a more usable format about two years ago. It’s in a GIS system now. We have sign records, and reliable information starting from approximately 2007 on signs.

Are the DataKind models available? Did any enhance your previous predictions? Not yet. They are still doing QA/QC and have lots of wrap-up to do. The exposure models are ready to be used for developing risk factors, however.

What is the state of the art in your department? We collect and maintain data, but do not have in-house statistician at this time.

Have bike incidents decreased on 2nd avenue since they put in the new bike lanes? We are close, but not quite ready, to evaluation; we usually want to wait one year for after data. Preliminarily: bicycling has gone up, as have collisions, but the overall rate is down. Vehicle collisions have gone down as well. We are also seeing pretty high compliance from drivers as well.

Near-misses or minor collisions – how can we track that data? Such as looking at social media reporting? We’re interested in all of the data that we can get our hands on. We do have 10-12,000 crashes that are actually happening, but are interested in every piece of information that we can get so we can be more practice.

Do you have traffic volume data for each mode? Bikes: Strava data comes from phones to track route and generate exposure information. We’re also trying to get counters for bikes on every bike facility.

Pedestrians: we are part of the national effort to count pedestrians? Quarterly we count at 50 intersections, and the downtown association does counts at intersections as well.

Vehicles: We track type and number of turning movements of vehicles at intersections. We have good volume and speed data for vehicles, often using tubes. We’re mainly concerned with volumes on arterial street networks, which we are using to create exposure models.

Police were still using paper tickets as of last year. When are they in transitioning over to more electronic records? In 2015, SPD just switched over to all-electronic citations. There are still limitations: officers are not standardized across their data collections, so lots of data cleanup needs to happen.

Contact Jim: Jim.curtain@seattle.gov 206-684-8874 Seattle.gov/visionzero Seattle.gov/transportation


Luke Swart, heyduwamish.org. They can use help from journalists, technologists, anyone interested in using this platform to protect our environment. Contact: luke@heyduwamish.org, lukeswart on Slack, or heyduwamish.org - click “Get involved”

Emma Brillhart is working on Feeding Seattle. Compilation of all the subsidized and free meal programs for homeless and low-income people throughout the city. Want to switch from Angular to React, and make it more usable for social workers and homeless shelters. Adding Better filtering, better features, better maps. General geographical data. Contract her at emma.brillhart@gmail.com, ebrillhart on Slack, ebrillhart at github.

Scott Pierce is working on WeThe!, a special engagement app allowing you to vote on legislation that is before your representatives at various levels of government. He’s looking for tech collaborations for a mobile app and the back end. Contact: scott@diamondstrategypartnership.com