Hackpads are smart collaborative documents. .

Steven Vance

863 days ago
Unfiled. Edited by Steven Vance 863 days ago
Steven V
  • Bikeplanner.org - no longer exists, allowed setting routing preferences set between Quickest, Safest and Flattest.  Also did routing involving Bikeshare bikes.  Based on OpenTripPlanner
 
  • Mapzen has instance of OSRM; anyone else offer instances (with open API) of OSRM?. Mapzen also offers a free and fast geocoder and autocomplete/suggestion tool called Pelias. 
 
900 days ago
E S Hacking Transpo Data!
Creating a new curriculum for transpo-tech professionals
(Working with GTFS, APIs, PEMS, etc.)
 
The People!
Danielle Dai (@danielledai)
Kendra Levine (@tranlib)
Kelly 
Melanie
Christine M Christine (@nspiregreenllc)
E S Jasmy
Alejandro
Ekta Ekta (ektadary)
E S Camille G.
Miles 
Andrew
Enrique
Dan Morgan (USDOT Chief Data Officer) (@dsmorgan77)
Kevine
Stephanie 
Christina
Jim Bunch
Holly
E S Ryan
Meghan Hade (@meghanhade)
Drew 
Austin
Anthony 
Anna Jose
Brian L Brian Laverty (@BrianLaverty4)
Chad Savage
Spenser R Spenser (@ionlyeatbeats)
Tiffany
Alberto
Justin G Justin (@justgrimes)
Rebecca W RW (@internetrebecca) From online 
 
Steven V Shoutout to Maptime. (Most people had heard of it.) Create something similar to this. I feel a lot of transportation apps come from computer programmers. Planners should be empowered to make this. 
 
E S Key Points:
  • Use data for teaching/learning transpo concepts.
  • Building confidence & learning the tools
  • When is it appropriate to use certain methods & knowing how data is collected (the assumptions)
 
Data Sources
E S
  • On time arrival information for planes
  • FAA Wildlife strike database
  • Surface
  • Geography
Steven V
  • OpenAddresses.io - 100 million address points for the world, to help you find places
  • GeoNames - 2 million points of interest (free, open)
Raphael D
  • Census TIGER
Spenser R
  • Pipeline Explosion
Raphael D
  • Freight
Steven V
  • Census of who has a US DOT (license) number
Raphael D
  • American Association of Railroads (AAR):  freight tonnage transported by category for USA, Mexico, and Canada (so  you can see the rise of transporting fracked oil from North Dakota to  Illinois and New York)
  • Road
E S
  • National Bridge Inventory
Steven V
  • Highway Performance Monitoring System (HPMS) - traffic volumes; ARNOLD All Roads Network of Links  Data will collect GIS roads
  • SWITRS - traffic safety for California -
Raphael D
  • MATOC.org
Holly K
Steven V
  • University-hosted crash databases for states: UC Berkeley's TIMS
Spenser R
  • Railroad Safety Data
Raphael D
  • Transit
E S
  • Public transit location recordings (not open :-( )
Raphael D
  • GTFS Real Time
E S
  • Fatality Recording System 
Steven V
  • Data.gov - can help you find which Federal Department has which dataset; this site also federates some city and state data portals. 
E S
  • DC RITIS - Regional Integrated Transportation Information System
Steven V
  • MABLE - Crosswalking Census geography
E S
Holly K
Steven V
  • APPS - WalkScore, Waze, Strava, RunKeeper - not data sources, but data interpreters
E S
  • Bikeshare systems (station locations, all trips)
  • VIA from Berkeley - transit operations to track performance of bus routes
Steven V
Justin G
  • Submit FOIA requests
  • write scrapers; extract from PDFs
E S
  • Open Data Portals from Socrata; CKAN
 
Rebecca W Data Standards
  • GTFS - The General Transit Feed Specification, originally developed by Google, contains static schedule information for transit agencies including: stop locations, route geometries and stop times;
...
900 days ago
Unfiled. Edited by Michæl Schade , Steven Vance , Anna Petrone , Peter Miller 900 days ago
Michæl S How Can we Use Data Collected from Transport to Catalyze Better Land Use?
Steven V Session 3, room 468
Alternative title: using data (from apps) gleaned from transport to better inform land use planning
 
Kim opening the session. She works for DCDOT and oversees bikeshare and bike parking. 
Better (different) land use to manage demand for transportation services/modes. 
Anna P In DC, there's a large residential section and large downtown section, means bikeshare can't meet demand. 
 
Steven V Q: Current land use and try to serve it better? Or projecting future land use and what services to deal with the future?
A: I think these apps were born out of making today's land use better connected. 
 
Spatial mismatch on where people are living and working. We should then know where to locate employment. 
Kim: Kind of. There's a reason that businesses co-locate. 
 
Q: I don't think this is a data problem. I think it's a communications and economics problem. 
A: We know where the buildings are, the people are. 
 
Q: Do you have a thesis?
A: In Chicago the thesis is that reverse commutes (from core to suburbs) and suburb to suburban core are poorly served by rail transit. 
 
Q: How are people visualizing the data? You only have a couple seconds to show and convince a city councilor about the topic or issue. 
A: Social capital, getting people to tell other people in their networks [about what?]
A: Probably the data we have but no one looks at it is land values. 
 
Q: Having a Rite Aid (single story cinder block building) across the street from the U Street Metro station is clearly a waste of space. 
 
Visualization idea: What existing tools, resources, data are out there?
  • VMT per capita (link?)
  • Perhaps maps of density, where people live and work
 
Q: City mandate that developments can only go where there's existing service?
A: An adequate public facilities ordinance in Arlington county, 5 TDM districts. Developments costs and fees. 
There are some municipalities that have TDM requirements. You can develop whatever you want in Cambridge, but you can't increase vehicle trips there. There can only be so many vehicle trips to that development. 
Grubhub advertising pizza. Can we look at their data to show where all the pizza is being delivered? That would be a key indicator that someone should open a pizza joint there. 
 
Q: I work with the National Household Travel Survey. What kind of questions could be asked to better serve land use planning?
A: Are people using ride-sharing, car-sharing? Can you conduct more multi-day surveys?
A: How low does the cost have to go? GPS and smartphones are those cheap devices that can do those multi-day surveys. There's a sampling issue if you rely on what devices people currently own. 
A: Mobile phone operators, INRIX. Then the problem is removing the personal data. 
 
Q: Buses know when and where people are boarding and deboarding. Uber knows where people are getting in and out of their cars. What clearinghouse should there be to collect and distribute this data?
A: Perhaps each area with an MPO. 
 
Q: I think there are a lot of forces to open data. And governments are on the right side of history for doing so. Do your local land use planning entities know where it is and how to use it? That's where local governments need help. Bikeshare is a great example, and people are playing with the data and visualize it. And it could lead to individuals to say, "look at what I've learned" after making this visualization and messing around the data. 
 
Q: The question of institutions is important. These data are available, but will the MPO actually use them? Boston: local MPO partnered with state DMV and got VMT broken down by 50 meter x 50 meter grids. They released it as a part of a visualization competition. So how do we move from the design competition to understanding the resource institutionally. What are the land use decisions that need to be made, and what are the data needs?
 
Q: Most people here in the room are data users and few are people who have data to provide. So many retailers want this data: who came into my shop? We had this event, what was the result?
 
Q: Not building up to the zoning limit, that's development left on the table. Form a more permissive zoning. Data could be used to show the development left on the table. Having data on where people are traveling (to certain types of stores) you could conceivably reduce VMT if you were to relax zoning in certain areas. Make the zoning responsive to the travel demand. 
 
Q: Grubhub. They don't actually sell food. They don't care if another pizza shop opens, they connect customers to food. They don't have that incentive to keep the data closely held, like someone like Rite-Aid might. Melbourne: carbon negative; adding services as infill in neighborhoods. 
 
Q: Bikeshare data knows about the trip but it doesn't know about the end use. Transaction data (credit card) is more telling about what I'm doing at places. 
A: Comparing mixed use with single use zones to see if the different zones can reduce overall VMT and relate that to bikeshare station density. 
 
Q: Land isn't changing but building is changing and the new user needs development approval. Residents want to know, is that going to impact me? And planners are required to say "yes" or "no". The data on what kind of trips that new building use will generate is really unknown (the data are weak). 
A: What data do you want, and how we can get it for you? What data are useful for land use planning? (make an audit and assessment of all the initatives and how precise they are)
 
Q: I think the game changer is not the data collection but the linkage of different sources of development. Get different parts of the elephant to patch together. We can't see that we all have different parts of the same animal. It seems we have the same problem with data. 
 
Q: Change the number of bikes, station locations, pay people to do the reverse commute?
A: DCDOT tried that but it didn't work, partially because the rewards were issued manually. Also, $1 isn't enough to get people to ride uphill. From another session: another reason this didn't work is because of the land use just doesn't support a different route/commute. 
A: IEEE conference analysis of Citibike NYC to push people to nearby stations instead of another route. 
 
Peter M Trip data sources
  • Crowd-sourced video analytics (a New York startup about 12 months ago?)
  • Mobile phone cell data from mobile operators, possibly aggregated by Inrix or Airsage.
  • Bluetooth ID sniffing
  • Data collection apps for smartphones
  • Automatic numberplate recognition
  • Telephone surveys
 
 
 
900 days ago
Unfiled. Edited by Ryan Rzepecki , Anna Petrone , Steven Vance , Spenser Rubin 900 days ago
Ryan R Smart Bike Share Data Tampa & Phoenix
 
Ryan@socialbicycles.com
 
 
Key Observations:
 
Steven V
  • Core assumptions (wanted to test this)
Ryan R
  • ~50% of signups via the mobile apps
  • Pay per use model seeing high adoption
  • Hub Location incentives are working
Anna P
  • Sobi bikes: 
  • GPS data from trips (bike lock has GPS unit)
  • similar to Call-A-Bike in Germany (text a number to unlock a bike)
  • can inform where to build bike lanes - and to prove that people use bike lanes
  • Can look up online where it is!
  • Has API but is currently not open
  • Self sufficient bike share
  • pricing - is a good value for active user, but way too expensive for casual users
Steven V
  • new model: fixed hourly rate; casual passes (gateway drug to habitual user) has outsold annual memberships (Tampa, PHX); annual member has 60 minutes of free time. 
 
Spenser R
  • Virtual / Geofenced hub locations
  • no-hub model is bad for the user and bad for the city
Steven V
  • no extra fees for locking up inside the hub, incentive to rent from a hub. small fee for locking up outside of the hub; the fee isn't 1:1 –$2 fee for locking outside of hub in Phoenix, and you get $1 or $1.50 if you bring it into a hub. I think we can make it more dynamic, perhaps based on distance or time spent outside of the hub. What if we surged, increased, or decreased the fee. Perhaps incentivize reverse trips (problem in D.C.). Paul DiMaio: we found that people didn't go out of their way to get a bike, we were essentially rewarding people (with gifts) to people who already had those reverse commutes. 
  • 30% of users are parking bikes out of hubs; 30% of users are bringing out-of-hub bikes back to hubs
Spenser R
  • rebalancing is the biggest operational burden, incentives unload some of that burden onto the user 
Steven V
  • How do you choose hubs? Follow some of the same siting guidelines as dock models, but also have our own feasibility model. Idea of what a station is: Rigid for dock-based. Each station is a $50,000 investment, so you don't want to have outposts that would really only need 5-10 bikes. So with SoBi you can have that with 5-10 bikes (pocket hub) and it's more feasible. 
 
Ryan Rzepecki: take this data and shape the cities. SoBi is new to the bikeshare industry. After a few weeks see these concepts proven out. Gonna issue data formally. Sneak peak to an industry audience, not all of this is public yet. 
 
We thought Phoenix would outperform Tampa but that wasn't the case. Tampa has more use on weekends which is probably more recreational use. 
 
Each bike has a cell connection itself. The bike receives your account number and PIN code and you validate on the bike with your PIN code each time. You can make a booking on the bike itself with an RFID. 
 
[Showed off dashboard data visualization for the users.]
 
Q: Is it possible to separate the software from the hardware and sell them individually, perhaps for e-bikes or scooters?
A: We used to have that, with an attachment for the bike. But bikeshare bikes need special features, like integrated lighting. Most companies are making sport bikes. Then we developed our own bike. We can manufacture at scale. We don’t operate a bikeshare system. We offer a platform for operators to make their own business choices. 
 
Phoenix had to wage severe political battles to get bike infrastructure installed. 
 
Q: How open is the data? Is it just for the buyer (operator)?
A: I think it will be best if it’s open. The clients obviously 
I don’t have the full rights to release the data publicly. Many cities, though, have open data policies and all of them have agreed that it’s better. 
 
People have been using the tech as we anticipated. 
 
Q: Can you see casual users buying memberships? What's the conversion?
A: It's only been a few months; don't know yet. 
 
Q: price point diference of student and regular annual?
A: $20 price break; $59 vs. $79. $5 per 60 minutes. 
 
Q: What happens if someone brings it onto their private property (theft)?
A: You can see the last user in our dashboard. There's going to be some amount of gaming the system. We can see the last location of the bike. 
 
Q: How often does GPS ping?
A: Was 30 seconds (as breadcrumb), but we're thinking of switching it to 5 seconds and uploading it every 2 minutes. 
 
Q: Open data, anonymity from operators and from users. 
A: That's something we're going to have to deal with. Taking baby steps to open that up. Can the user see this quality of data on their own profile? Yes. We also have a social network. 
 
Q: Are you doing any user surveys?
A: Not yet, will soon. Our interaction with the bike is different than the dock-based models. More of a UX basis, where people have struggles with the app. 
Q: Traditional bike share we're trying to answer if that shifts people out of auto trips (mode choice). So I'm wondering if SoBi-style bikeshare can shift more. 
A: Yeah, and I think it can be improved if it's integrated, with transit RFID cards, or other apps, so the user can see "there's an Uber bnearby and there's a SoBi bike nearby". 
Paul: CaBi does a bi-annual survey: 54% go to Metro station. 
 
Q: Have you compared bike speed between annual and casual members? People who pay per minute, maybe they're increasing their speed. 
A: Well, we have that data!
 
900 days ago
Unfiled. Edited by Michæl Schade , Anthony Gallo , Steven Vance 900 days ago
Michæl S Smart Bikeshare Data
 
 
Steven V This hackpad has been moved to Smart Bike Share Data Tampa Phoenix
 
 
 

Contact Support



Please check out our How-to Guide and FAQ first to see if your question is already answered! :)

If you have a feature request, please add it to this pad. Thanks!


Log in