Thursday, June 23, 2005
The RobCast: RealityFlythrough w/Neil McCurdy
I recently saw an article on Roland Piquepaille's web site about a project called RealityFlythrough. The project aligned so well with where I see pervasive cameras taking us that I just had to talk to Neil McCurdy, whose doctoral candidate project this was.
Fortunately, Neil was gracious enough to spend 45 minutes to talk with me via Skype about what RealityFlythrough is all about, and to touch a bit on where it may be taking us.
Here's a link to the audio interview (~45 mins, ~21MB). Be warned, I noticed at least one drop out at the very end, and there may be more. If you find a significant one, let me know and I'll see if I can recover it from the raw audio.
Here are links to some of the things discussed in the interview:
- Neil's Web Site
- The link to the RSS feed for his site
- Neil mentions the Human Pacman game developed at the National University of Singapore (hmm, I'm getting a DNS error on my original link to the project, so if you find it broken, try this link instead and scroll down)
- I mentioned an augmented reality project concerned with projecting virtual objects into real space. I was probably talking about this, although here I mention something that seems in the ballpark as well (the destination link on this page also goes to the Singapore server that I can't see at the moment).
- Neil mentioned an Intel research project called Place Lab which is an effort to provide location data both inside and outside, and in a privacy-aware manner. Very interesting.
As an aside, Neil brought up privacy as a potential hurdle to enabling ubiquitous cameras to perform interesting and useful functions. Certainly we'll be running in to this issue of the creeping loss of what we have traditionally seen as "privacy". I think this warrants a series of discussions all on its own. I'd love to hear any opinions out there.
And while you're giving me your opinions, I'd also love to hear about any other cool, forward-looking, potentially tranformative projects going on out there, and also where you think they might be taking us in the future. Feel free to leave a comment below, or drop me an email.
Monday, June 20, 2005
The Internet in an Emergency
[NOTE: This is the first article that's come out of an experiment I'm doing using Wiki's as a way to formulate thoughts. This one came out fully formed, so it's kind of a cheat, but the Wiki space will hopefully help thoughts grow more organically rather than having to come out complete. Check it out if that seems interesting.]
I've been doing some work on an information system for my neighborhood, so issues relating to information flow in a geographic area have been floating around my head a bit.
It occurred to me that if there is an emergency in a neighborhood or city, that restoring Internet access is increasingly critical.
Given that people are already using the Internet as a primary means of communication, with VoIP capturing an increasing number of users every day, restoring Internet to an effective area has become at least as important as restoring phone service.
Given it's ability to coordinate groups and disseminate information, it is becoming as important as restoring power (though there is a ways to go yet), and it's importance will only increase over time.
I wonder if our neighborhood emergency plan (or the city plan covering our neighborhood) includes an emphasis on getting some semblance of the Internet back up after a disaster. I'd guess not given the wide range of providers and mechanisms involved. For example, do you bring up cable first? DSL? Cellular?
Probably it makes better sense to create a few satellite uplinks where people could bring their increasingly portable computers and connect in pre-determined locations.
Does your community include a plan for restoring communications in an emergency? Does it include the Internet?
Let me know. I'd be interested to hear about it, how it works, and how it got in to the plan.
Sunday, June 19, 2005
Neighborhood Information System
Here's an idea that's been circling in the back of my mind for the last couple of years: a Neighborhood Information System. Basically, this would be a place where all members of a community could contribute information to a central location for the benefit of other people in the neighborhood.
I have some pretty ambitious plans for what such a system might be like, and what it could do...which means it will probably never happen.
The GOOD news (depending on who you are, I suppose) is that it occurred to me that there are plenty of useful tools out there right now to do a pretty good first pass at such a system without having to build it from scratch. Wikis, for example are a good place to host information that is relatively static, like neighborhood history. Blogs are a good way of providing news, and message boards are a great place to share dynamic information, among other things.
My short term vision is that the best of breed open source versions of these tools would be combined into one system for the primary purpose of providing a single login for all of the tools. This would then be released as open source for any other neighborhoods that wanted to put it in to place, and potentially add their own improvements.
The driving force, however, is to allow ANYONE to contribute to ALL aspects of the system. This means add/change historical information, create news items, and create a new topic in the message board. I believe this system will work best if there is no centralized control (though I DO think it's appropriate to have oversight committees for such things as taxonomy, design, and contribution guidelines). Instances where a single person, or group of individuals should have absolute control over some portion of the system should be exceedingly rare.
It is in this spirit that I've started a Wiki for the Laurelhurst neighborhood in Portland, Oregon. This is a test set up on a hosted wiki service that uses ads for revenue. Ultimately the goal would be to get on a system that was not ad supported and ran an integrated system, but at this point I'm thinking baby steps.
If you live in the Laurelhurst neighborhood, feel free to drop by, sign up and start adding stuff. If you are interested in the concept and think it would be interesting, drop me a comment even if you don't live in my neighborhood. As I said previously, I see this as being something that every neighborhood could use for free (though you may have to provide your own hosting) so there's no reason it has to just be about where I live.
I've looked around to find if such a system already exists and didn't come up with much luck. But if YOU know of such a project in progress, leave a comment with a link. I'd much rather not rebuild the wheel if I didn't have to.
Wednesday, June 15, 2005
Underground Automated Highway System (UAHS)
John Smart recently included a vision paper on a possible future mode of transit in the most recent issue of the Accelerating Times newsletter. Note that this issue references the availability of the audio portion of Peter Thiel's presentation on Virtual Money at last year's Accelerating Change 2004 conference, where yours truly can probably (I haven't listened to the audio yet) be heard asking the one question I worked up enough nerve to ask during the conference.
He provides a vision for a form of transportation that would relieve the most crowded cities of their surface traffic burden. He envisions a network of underground tunnels in which vehicles can travel in a completely automated fashion. The details of Mr. Smart's vision will not be explicitly repeated here. His short paper on the subject is very consumable, so interested persons should get the PDF here.
Mr. Smart's paper was not intended to be a thorough treatise on the subject of underground travel, but rather to lay out a vision of a
possible future that he believes is likely to unfold.
Similarly, my thoughts are not intended to be a result of a thorough survey of the available literature. They just happened along the way, and I'd guess that these (and probably more) questions would pop up in the brain of the average reader of this paper.
While I am yet un-opinionated ;) on the topic, I did find that it left me with many questions. What follows (in the tradition of my first PRT entry, which seems to have generated some conversation) are my notes and questions as taken during my reading of the paper.
I should note that any errors that show up in quotes I pulled below are probably mine. As for the errors outside the quotes...well I'm still looking for the person that put those there ;)
In what ways does the Underground Automated Highway System increase the "...capacity of goods and transport in our largest and highest
density cities..." that the automation of existing roadways would not?
Would it be due to additional automation that is unlikely on existing roads?
Would it be because of the supplemental real estate devoted to transportation beyond the existing infrastructure?
Or does it simply by the addition of usable lanes?
How does this system increase peak traffic capacity beyond the additional lanes it provides, that are unlikely for on-surface, or
above-surface transportation alternatives?
How does it do so by an order of magnitude?
UAHS System Architecture
Is this system strictly oriented around major hubs, or does it work for more distributed trip patterns?
Where would entry/exit points be on the UAHS, and is there potential for congestion at these points?
If you were going to have a fully-controlled environment, could you regulate the flow of air so that a one-way tunnel had air flowing through it in the same direction, reducing drag?
If an "automation test lane" is required, how much lead space is required from an exit point (from a regular highway) for a vehicle to establish its autonimity? Would the driver have to cross existing lanes from, say, a regular freeway onramp, to the far left lane? Obviously this would create its own problems if the freeway was under heavy traffic load. How does this lead space (to establish the autonomous nature of a vehicle) compare to the average commute distance? How much of the original commute waiting time might still be experienced?
Does a megalot become a location where many people would depart from the UAHS, and thus become a bottleneck?
Does a UAHS suffer from congestion for its feeder networks (roads)?
What is the trip length that UAHS would be designed for? The way it's talked about makes it sound like you'd be traveling in it for a while.
If it is true that cities saturate as described, might it be the case that UAHS' order of magnitude more supply might be overkill should it be attainable?
Future Traffic Patterns
Do we believe that traffic patterns will get more focused on fewer destinations over time, or more distributed over a wider range of
origination and destination points?
"Nearly all population growth for the foreseeable future will occur in urban areas...". Does this necessitate that they will happen
primarily in the urban core, or will they be distributed around urban/suburban areas of a city.
Crikey! 8 lanes each direction? Will future demand be that high? What does the world of transportation demand look like 100 years out?
Wouldn't increasing real estate prices in the urban core tend to drive businesses into the suburbs, thus decreasing the number of travelers to that area? Is that effect measurable? Does migration away from the suburbs create new focal destinations, or do they become relatively distributed? Presumably the latter case would be harder to service with something like a UAHS.
How are existing non-car alternatives to a UAHS inferior? What about those on the horizon?
Surely automated aerial transportation systems are not the only viable alternative to a UAHS? What are the design requirements driving this as the selection set?
Issues with ridership of transit alternatives may be due to the format of existing forms of transit: Linear, multi-stop trips, with limited flexibility in departure time and destination. Perhaps alternative transit forms like PRT or dual-mode transit would increase ridership with less financial impact.
If there is evidence that the cost ratio between surface vs. elevated vs. underground transportation costs is improving for underground transit modes (from 1:3:6 to 1:2:4.5), does the evidence support that elevated transportation costs will also continue to gain? If elevated transit took the form of small guideways (e.g. PRT), could that improve the costs to the extent that they are BELOW costs for surface transportation?
If the same dollars that it took to create UAHSs were, instead, put to improving existing road systems and dedicating lanes for automation where possible, what kind of effect might it have? Would it be enough to offset the benefits of an underground system? What if the ground-level automation included significant numbers of busses? What if this money was diverted into other transit options such as LRT, PRT, or dual-mode transit? How does one establish the average ROI for each transportation alternative?
Would the costs involved in underground construction truly become advantageous over on-surface or above-surface transportation?
Costs per cubic foot quoted ($1.50) for "mining and lining" should be reduced to costs per square foot to be more directly compared to other forms of transportation, and all supporting infrastructure costs should be added to this to provide an apples-to-apples comparison of finished, working surface area.
Is the disposal cost of tunnel ejecta included in the per cubic foot dollar figure? What is included in this figure?
What would the average cross-sectional area of a tunnel be in this system, and would that account for one-way or two-way traffic?
At $1.50/cubic foot, assuming a square tunnel 16ft on a side (which is probably an order of magnitude too small), the linear cost of digging is $384/ft. At 100ft on a side, it's $15,000 per foot. That's $2,027,520 - $79,200,000 per mile before you lay any infrastructure (e.g. roads, ventilation systems, etc).
Lane widths in the US appear to vary between 9' and 12'. Using 10' widths to make calculations easier, a 4-lane road would be about 40' wide. Assuming that shoulders would be 5' each (again for ease of calculation, 4' appears common) to provide enough ceiling clearance for vehicles traveling in the outer lanes (would be interesting to see if normal long-haul vehicles would be able to travel in these lanes, but not critical for a number of reasons such as the fact that such vehicles may not meet emissions standards enough to use a UAHS, may be redesigned to have a lower maximum height in order to use a UAHS, or may be required to travel in the middle lanes), the total width is 50'. Assuming a circular bore, with a 50' diameter, the cross sectional area would be (pi*(25)^2)), or ~1963 sq.ft. At $1.50 per cubic foot to tunnel, and 5280 feet/mile, the cost/mile is: $15,550,884 for just the tunneling.
The low cost to convert existing roadways to an Automated Highway System (AHS) ($10,000/mile) seems largely irrelevant to the UAHS argument as all savings is lost in the initial construction costs of the tunnel and roadway.
What would the average cost of a "megalot" be?
Emergency Systems and Safety
How would emergency access to such an infrastructure work?
How do various types of emergency plans (that is, for various types of catastrophic failure) compare? Also, how does the probable incident rate compare with each transportation mode?
AHS's are probably safer in general, but, to the extent that non-automated traffic runs in the same area, this safety may be
somewhat compromised. Also, there is an open question as to whether significant numbers of individuals (1 in 100 perhaps) would break out of the automatic control to try to game the system, or just plain speed. Similarly, the extent to which an automated system can handle unexpected obstacles will heavily determine how much emergency response is needed (and, potentially, the limit of wide-spread
adoption of the system). If an animal wanders on to the automated roadway, for example, how does the system react? All in all, it is
not yet apparent that any reduction in law enforcement/emergency response could actually be experienced (although it does seem likely).
Granted that automobile accidents do account for a significant number of deaths each year, but if there is no public outcry, even in the face of general knowledge of the danger, then it would seem that safety cannot be a primary motivator of change in that other safety issues will always gain priority for the dollars.
By "...eliminating visual blight and noise pollution of freeways.." is Mr. Smart suggesting that existing infrastructures will not just be supplemented, but replaced? If so, how will vehicles not presently equipped to run in an automated mode make trips that would otherwise have required a freeway?
"Productivity loss" has always seemed like a weak reason to improve transportation infrastructure. Is this productivity loss to business (in which case it would seem a result of poor worker planning than anything else; if we're talking about people arriving late on the job), or to the individual (in which case it would be tough to assign a dollar figure to the time of non-working individuals)?
How does a digging rate of 20 ft/hour compare to other transportation alternatives when all prep-for-build requirements are taken into
Are there rights-of-way that need to be purchased/otherwise accounted for for underground construction? Can underground construction be built directly from point-to-point if there is no engineering reason not to? Said differently, are there places that a Tunnel Boring Machine can't go because of political/social reasons that it might otherwise be able to navigate just fine (e.g. below a residential area)?
Among the stakeholders should be those who have a stake in alternative forms of transportation. The automobile industry. And potential system component manufacturers.
What is the likely timeline for UAVs? Likely the biggest hurdle would be social as even automated cars are likely to be able to be
overridden by a human operator in case of failure. This will probably be an issue more of perception than anything else.
How do we prevent - or is it even undesirable - reclaimed roads, should there be any, from becoming more buildings as they would seem
likely to do if real estate continues to become rarer?
It would seem that other possible futures include: the emergence of attractive transit options such as dual-mode and/or PRT that have a good deal of the positive effects (and some added ones, like less reliance on owning a car) of a UAHS with potentially smaller costs; telepresence and telecommuting taking hold requiring fewer people to go to point destinations (like the urban core) simultaneously; and the possibility that trip destinations become more diffuse rather than more concentrated, making the value proposition of UAHS less appealing.
Sony Data Tiles
A very cool integration of physical objects and computing interfaces, and similarly cool integration of conceptual objects with eachother.
Basically a screen you can put various tiles on to to interact with the computer. Much cooler than I can make it sound.
sony revolution (the newest crazy technology from sony - nifty) - video
Monday, June 06, 2005
The RobCast: GeoWiki
Here's an idea(~16 mins) I had about how geographical information can be maintained, wiki-style, by everyone as a way to jump-start the Internet's connection to the physical world, and to ensure no one company owns the data.
Thursday, June 02, 2005
The RobCast: Wikipedia, Time Lines, and Our Stories
This is another "commutecast" (10 min), which largely consists of me huffing and puffing noisily after I get out of the car and walk the few blocks to work.
This time, an older idea of mine (one of my "mental whittling" projects) - The Web of History - given new inspiration by Wikipedia and a recent article in New Scientist, and a new idea about collecting human stories on a global level to add depth, perspective, and color to encyclopedia, and other fact-oriented mediums.