Exploring the Future Potential of AI-Enabled Smartphone Processors
GIS-T April 2010 findings
1. GIS-T Conference
CONFERENCE : GIS-T
LOCATION: Charleston, WV
DATE: April 12-15, 2010
PARTICIPANTS: Steve Lewis, Patricia Solano, Richard Grady, Michael Terner, Todd Barr
CONFERENCE NOTES
GIS-T 2010 was hosted in Charleston WV April 12th-April 15th, 2010. The TFTN-Strategic Planning Group hosted a panel
discussion of 8 speakers, to discuss what is TFTN and what does it mean to me. The panel discussion took place on
April 13th.
Speakers- Federal Perspective
• Steve Lewis - DOT
• Randy Fusaro - Census
• Ronald Vaughn – DOT
• Dan Wider – VDOT, but representing NSGIC
Speakers – State Perspective
• Tamela Lang - Colorado DOT
• Skip Turner – NavTeq
• David Blackston – Ohio DOT
• Melanie Rippon Seigler – Virginia DOT
April 12, 2010
The panel met during the afternoon break to discuss their thoughts and give a general overview of their presentations.
The following are notes harvested from that discussion.
• US DOT culmination of FGDC one transportation data model
• FHWA how can HPMS support the TFTN
• Tammy – more work for DOT outside of our business line, recommend to use HPMS. Common voice, open
street did not know HPMS. Common geometry used for different purposes. Using private products use private
partnerships.
• Randy – Federal requirements.
• Parker – NavTeq- they can provide a routable dataset quick. Then the government can take over an add to it.
• Randy –
• Dan Widner identify basic needs build it once use it many times
April 13, 2010
The Panel Discussion began at 8:00am. Patricia Solano and Richard Grady have a brief overview, that was followed by
the presenters.
PRINTED: 10/20/2010 PAGE 1 OF 7
2. GIS-T Conference
The following are notes harvested from the first panel discussion, which focused on the federal perspective.
Steve Lewis Points
• Feds shouldn’t tell YOU what to do. We want to hear from you on how to do it right.
• All modes of transit eventually
• Routing, LRS, addressable à Common geometry
– Private sector adds routing?
– Census adds addressing?
Ronald Vaughn’s Points
• Nation trans data source with surface system attributes
• Feds: coordinate/reporting reqs; States: data authority for collection; Locals: acquires and aggregates; Private:
support & technical services
• States should take an enterprise approach: GIS = ARRA; Traffic; Accidents
• HPMS can influence states via reporting requirements; would like to assist states, if possible
Randy Fusaro’s Points
• Core features – NOT everything to everybody
• Aggregation from the bottom-up
• Planning begins with ensuring Federal agency road reqs are met; then moves on to cover broader “national”
requirements
• State & Local: populate and maintain the roads
• Private sector: YES, but the “features” should be in the public domain – value add “linked” to public domain
• Other stakeholder: NGOs & academia
Dan Widner
• NSGIC wants to see people get together in a coordinated fashion. “Coordinated approach”
• Leverage the VGI/crowdsourcing activities
• Attributes: begin with the bare minimum attributes – addresses
• We need to identify everyone’s “business needs”. Not about ownership, identify what people need.
The following are questions asked/comments made by the audience and the panelists responses after the first panel
finished
• Volunteered Geographic Information (VGI), are there any “active plans” from panelists? Steve: reaching out to
OSM. Randy: USGS had an external VGI conference, Census participated. Open questions “how do we control
and incorporate”.
• Gene Parcher, USGS: How do envision the states owning the standards? 50 potential standards? Ron: Feds
would help the states arrive at a common standard. Reporting requirements can drive that. Steve: it doesn’t work
for the Feds to dictate, need input on what states should provide.
• Is basic routability part of the TFTN vision? Steve: we would like to see a complete, routable data set in the public
domain. Randy: it would be difficult for the Feds to do it well. Especially the maintainability part. Bruce/CSI’s
followup question: Distinction between “basic topology” (connectivity) vs. “fully routable” (turn restrictions)? Ron:
Yes, we’re looking at basic topology. Steve/Randy: YES, basic topology.
• CO-DOT: Concerns about VGI. Concerned about losing authoritative source and trust. The crowd isn’t already
right. Dan: what is the definition of “authoritative source”. Different needs – finding a restaurant, get close;
sending an ambulance, no margin for error. Nebraska: agrees with Lou/Co. What’s true and trustable? Randy:
PRINTED: 10/20/2010 PAGE 2 OF 7
3. GIS-T Conference
trust, but verify. How do you verify is the big question. Dan: we can’t be asleep – these technologies are catching
on and we can’t be left behind.
• Answer to VGI. There are editors from the public contributions. How soon is the data checked? A formal process
for checking.
• Kentucky Transportation: The public is willing to live with an imperfect world. We need to adapt to a faster
turnover. Capture metadata on who collected the data. You can see trends on who provides quality.
• VA-DOT: What does routable mean? What does roadway mean? What does network mean? What about a
higher level standard like an ISO process? More standardized definitions? Steve: no such effort underway. But
the standards could be part of this process. Randy: first need to identify the core features and attributes, then we
move to the standard. Dan: we are behind, other are leading the way. We need to set the lead.
Panel Discussion 2, the State Perspective
Tamela Lang
• What does it mean to me? Potentially more work without resources? Why can’t we use something that’s already
been done? Why can’t we partner the private sector and work deals with them?
• There’s less duplication in state govt. than people thought
• Private sector: “data enhancers”
• Public: error reporters
• CO has statewide roads, we meet our HPMS requirements. HPMS is the starting point; build our LRS on top of
that. Add bridges and accidents and addresses ON TOP of that.
• Clarify – Transportation DATA for the Nation
Skip Turner
• TFTN is a consistent base map with common structure used across applications. All parts of the country –
rurals have less quality in commercial products.
• Cross jurisdictional – city-to-city; state-to-state; US-Canada
• Public-private partnership models are workable. We will need to continue to charge for the special sauce. We
have field geographers and do validation.
• Willing to work with you to put the “geometry into the public domain”.
Melanie Rippon Seigler
• Melanie/VDOT: Federal, state, local “collaborative”
• CL’s are the basis for LRS and routable
• We need to think of multi-modal going forward
• Feds should coordinate the standards development by states
• States should educate and support local govts.
• Other stakeholders, Mass Transit organizations
• In VA state provides good orthos to support local govts
• Keep the attributes simple
• Census population can help prioritize data colllection.
David Blackstone
• Create it once, use it a bunch. Uses of centerlines have expanded over the years. Accidents, emergency
response. It’s not just mileposts
• Federal requirements = once/year; doesn’t work for local govt. need a more transactional model
• Avoid a “top down” implementation
• E911 can be a key ally
• Other stakeholders US Postal Service
PRINTED: 10/20/2010 PAGE 3 OF 7
4. GIS-T Conference
The following are questions asked/comments made by the audience and the panelists responses after the second panel
finished
• Steve Lewis: States, would you submit a “complete road network”? Should we change the requirement? Melanie:
we have it, we would provide it, but it needs to be a requirement that create a business need. Blackstone: NO,
the attributes are extensive and need to be defined first. “Roughness tests” are expensive. Relax some attributes.
Subsets, we could do it.
• GA-DOT: Mandates get resources assigned to them. Can’t do something “for the good of the public”. How does
TFTN work if there’s not strong internal state coordination? Blackstone: Be creative in finding the money. Tie-in
to safety. Census spent $24M on improvements in OH. Spending $’s to locals is popular and politically workable.
• VA: E911 and safety funds are key participants; homeland security funding.
• LA-DOT: Emerging issue is to tie-in DHS and emergency managers.
• LA-DOT: Google is now using NON commercial data. But the data seems degraded. What is the definition of
acceptably accurate? Emergency managers have accuracy needs. Blackstone: The commercial providers and
Google are asking for our data. Skip: New products LiDAR + panoramic cameras are being driven now. 3 cities
will be available this summer. New York, Chicago and Los Angeles
• Steve Lewis: Skip and Virginia, you have a partnership for HPMS? Is this a prototype for TFTN. Skip: Not
identical products, but two parallel products that share geometry. VDOT doing daily updates. In MA, we did lots
of QA that improved the geocode success >90%.
These presentations can be found it http://www.transportationresearch.gov/TFTN under TFTN documents GIS-T 2010.
TFTN Breakout Session Tidbits:
• Bruce Spear of Cambridge Systems: Don’t necessarily need all HPMS attributes down to local level
• Mark from FHWA asked about examples of TFTN-like approaches in other countries
• Jean Archer of USGS: Is US unique? What is HPMS geometry
• Ron: HPMS geometry is the state’s base network of road geometry that attributes can be attached to
• Steve: What is the relationship of State GIS Coordinators with State DOT GIS Managers?
• Randy: Leverage federal resources to create common core data set; start with roads; look at funding
availability; TIGER is a good place to start; you can hook-on data to TIGER; Fairfax County example
mentioned
• Bruce: If a state DOT had better data, would Census use it?
Emerging Issues Panel
• Jim Mitchell/ Louisiana DOT: The National Map and a National Transportation Layer – is it part of the solution,
or part of the problem? Where does the money come from? What are transportation networks used for (e.g.
emergency management), and what do they need?
- Put it in a map
- Geocoding
- Analysis
• Costi/DHS IICD: HSIP data used by DHS and federal HIFLD community; states have access through HSIN,
using iCAV
• Data should be “Accurate, Authoritative, Accessible, Actionable, and Affordable” (the Five A’s)
• Bruce Spear: Fundamentally, we need geometry and network topology, with addresses.
• Randy: TIGER was tested against survey quality points, and has feature-level metadata
• Mark from TeleAtlas: Update frequency may be important; TA offers quarterly updates, but some customers
have trouble absorbing a new dataset every 3 months.
• Randy: TIGER updating depends on funding to be determined by Congress via budget process
PRINTED: 10/20/2010 PAGE 4 OF 7
5. GIS-T Conference
POST CONFERENCE COMMUNICATIONS
On April 16, 2010 Richard Grady sent an email with his post conference analysis
Rich Grady's Follow up April 16, 2010
Hello Patricia and Steve,
I thought it might be a good idea to share some quick thoughts based on our recent GIS-T experience, while it was still
fresh in our minds. I defer to the two of you on overall marching orders and priorities, but as input for you to consider,
Michael and I have compared notes and impressions, and would like to suggest the following:
• Supply notes to Patricia for combining with what she captured, as the “GIS-T TFTN Workshop Record”; these will
be more of a running record than a distilled summary (aiming for end of next week, 4/23, to Koniag).
• Distill our notes into “GIS-T TFTN Findings & Recommendations” summary presentation – certain things jumped
out at us, and we want to make sure we captured the most important takeaways in a cohesive way; we will take the recap
slides that Michael produced in real-time, and refine and build on these in order to produce the envisioned summary
presentation – it will be something we can build on and refine as potential presentation material for future events and
webinars (we’ll aim to have it ready in draft form by the end of next week, Friday, 4/23, for full team review and comment);
a document version of it could also be produced, as a “strawman” position paper, but we should talk about this, first.
• Draft an outline for our final report, which we believe may evolve over the course of the project, but can at least
identify areas where Koniag and AppGeo can apply effort to optimize our overall budget and split the work most
effectively; from experience doing many Strategic Planning projects, we find that a good outline, done early in the
process, can help guide efforts, and balance the use of time.
• Aim for the end of the first week of June, perhaps June 3rd-4th (Thurs-Fri) for follow-up meetings in DC; we
suggest interviewing some stakeholders (e.g. HPMS program owners, TIGER program owners) as well as have team-time
to prep for the ESRI UC; I’m sure we’ll have regular communications between now and then on both of these subjects.
• Coordinate on other action items (e.g. Steering Committee, NARC Webinar, Other Events, Quarterly Report).
So, that’s what’s on our minds up here in Boston, where “once the embattled farmers stood, and fired the shot heard
round the world!” Apropos for both TFTN and Patriot’s Day!
Happy Friday,
Rich
On Monday April 19th, 2010 Steve sent out his thoughts on the Workshop
PRINTED: 10/20/2010 PAGE 5 OF 7
6. GIS-T Conference
Dan and Gene,
My thoughts on last week’s GIS-T conference:
The Good: The plenary session and the breakout session generated good discussion with more audience
participation that I anticipated. Even more telling, I was approached by a couple of State DOTs during the
breaks, and they were both very enthusiastic about TFTN.
The Bad: Some of the State DOTs seemed lukewarm on TFTN, mainly because they feared it would become
yet another unfunded mandate. Although not TFTN specific, I also learned some things from the survey that
each State DOT was asked to fill out. From the survey, only 65% of the State DOTs responded that they work
closely with their State GIS Coordinators. That leaves 35% that work “somewhat” or “not at all” with their
State GIS Coordinators. When I broached this subject in the breakout session, one State DOT rep said that
involving his State GIS Coordinator would “at best make no difference, but would more than likely make things
worse.”
The Ugly: At the wrap-up session on Wednesday afternoon, when the subject of the TFTN panel came up, one
highly placed member of the GIS-T Task Force and Planning Committee stood up, looked me in the eye, and
announced to the room “if I have to sit through one more TFTN panel discussion, I’ll slit my wrists.” This
particular gentlemen is a perfect example of the stove pipes that we have to overcome.
Let me know if you have questions!
Steve
Michael Terner Reply Monday April 19,2010
Steve:
PRINTED: 10/20/2010 PAGE 6 OF 7
7. GIS-T Conference
Great summary. I would like to emphasize that we felt like a lot of very valuable information came out during
the “good” part of the summary. Indeed, we have many new ideas on how a TFTN could be crafted and how
the roles of different stakeholder might play out. As per Rich’s “GIS-T follow-up” email from last Friday, this is
what comes next for us: sorting through what we heard, and figuring out what it means and then figuring out
how to spin it for a broader audience.
Finally, regarding the “ugly” comment, I can see where he’s coming from based on the good amounts of stage
time that this effort got at GIS-T. At the same time, I think it’s up to us to prove him wrong. I would hope that
by the next time he hears from us we will have developed a much different and more finely targeted message.
Indeed, we’re only at the “What do you want? How might it work?” stage of our planning. As we start forming
more concrete proposals and/or recommendations, the message will be a lot different – and hopefully more
compelling (even to the Doubting Thomas’s of this world).
MT
PRINTED: 10/20/2010 PAGE 7 OF 7