NYTimes.com: Guided by Computers and Sensors, a Smooth Ride at 60 Miles Per Hour
andre...@gmail.com
|
|
||||||||||||||||||||
Dave Petrie
--
You received this message because you are subscribed to the Google Groups "transport-innovators" group.
To post to this group, send email to transport-...@googlegroups.com.
To unsubscribe from this group, send email to transport-innova...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/transport-innovators?hl=en.
No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 9.0.862 / Virus Database: 271.1.1/3186 - Release Date: 10/08/10 23:34:00
Brad Templeton
Pretty limited until I get a chance to see the cars later this week.
While I have outlined the many benefits of robocars here and at
robocars.com, I focus the most on non-capacity issues. But there are
quite a number of factors where robocars increase road capacity
greatly, especially once they come to dominate the road. And closer
drafting is actually among the least of them!
Some of the other ones include:
a) The promotion of the use of half-width cars that can go 2 to a
lane, including 2 person "face to face" cars.
b) Better regulation of traffic volumes to not exceed road capacity,
resulting in major reductions in congestion.
c) Also not having accidents, rubbernecking and engaging in less
rational human behaviours that cause congestion even with no accident.
d) Perfect synchrony with traffic lights
e) Automatic load balancing of road networks to make use of large
amounts of unused capacity
f) Dynamic rerouting of roads (ie. make 7/8 of all lanes go inward in
morning, outward in afternoon) for 1.75x increase in surface capacity
g) Reliable elimination of street parking at high load times, for a
major increase of capacity on narrower roads. (Consider a 40' wide
street that today has 2 lanes of parked cars and 1 10' wide lane each
way, changed to 7 lanes all in the same direction for 4' wide 1-2
person cars, with no congestion or stopping for lights and stop signs
that's as much as 10x increase in capacity on that one street.)
h) Dynamic robo-bus service where 20 pod robocars converge in a small
lot at the same time, everybody walks onto a bus which then crosses
the high-load area to a spot where 20 other single person pods wait to
disperse the occupants. While not likely to be needed much, this
allows immense increases in capacity.
All of this without changing headways for drafting.
However, many of these are for a future world with a high density of
robocars. Let's not lose sight of the importance of Google's
announcement, which is a fleet of cars that are doing remarkable
things *today*. Driving 1,000 miles of city streets and highways, in
traffic, without a human even thinking an intervention might be
needed. While many have called my robocar predictions optimistic,
this has happened even sooner than I expected.
On Oct 9, 1:14 pm, "Dave Petrie" <DavePet...@comcast.net> wrote:
> E-Mail ThisThe automation of driver functions has the potential of being safer than the average driver. But we should not lose sight of the more global objectives: Increasing throughput on our existing highways, reducing dependence on foreign oil, reducing GHG. This means higher density is required to achieve the necessary throughput.
> The AHS* experiment in San Diego tried to do this, with cars having headways of a few feet. But due to fail-safe requirements, we will Never Get There. That is, we cannot afford to have a haystack wreck several times a month somewhere on our national freeway network. To achieve the fault tolerance to, say, the level of a modern airliners (FBW) flight control system (my specialty at Boeing) is not possible in the environment of the AHS concept. (too many places for faults to sneak into the system)
>
> *AHS- Automated Highway System
>
> ----- Original Message -----
> From: andress....@gmail.com
> To: transport-...@googlegroups.com
> Sent: Saturday, October 09, 2010 12:06 PM
> Subject: [t-i] NYTimes.com: Guided by Computers and Sensors, a Smooth Ride at 60 Miles Per Hour
>
> SCIENCE | October 10, 2010
> Smarter Than You Think: Guided by Computers and Sensors, a Smooth Ride at 60 Miles Per Hour
> By JOHN MARKOFF
> Designers of self-driving cars have made advances in detecting pedestrians, street lights and lane markers, as well as in resolving conflicting sensor data.
>
> Copyright 2010 The New York Times Company | Privacy Policy
>
> You received this message because you are subscribed to the Google Groups "transport-innovators" group.
> To post to this group, send email to transport-...@googlegroups.com.
> To unsubscribe from this group, send email to transport-innova...@googlegroups.com.
> For more options, visit this group athttp://groups.google.com/group/transport-innovators?hl=en.
>
Brad Templeton
Should note that the link above is actually to a sidebar on the
article, the main article is at https://www.nytimes.com/2010/10/10/science/10google.html?_r=3&ref=technology
There is also a video and some graphics.
On Oct 9, 12:06 pm, andress....@gmail.com wrote:
> This page was sent to you by: andress....@gmail.com.
> SCIENCE | October 10, 2010
> Smarter Than You Think: Guided by Computers and Sensors, a Smooth Ride at 60 Miles Per Hour
> By JOHN MARKOFF
> Designers of self-driving cars have made advances in detecting pedestrians, street lights and lane markers, as well as in resolving conflicting sensor data.
>
>
> ----------------------------------------------------------
>
> ABOUT THIS E-MAIL
> This e-mail was sent to you by a friend through NYTimes.com's E-mail This Article service. For general information about NYTimes.com, write to h...@nytimes.com.
>
> NYTimes.com 620 Eighth Avenue New York, NY 10018
Jay Andress
--
Richard Gronning
If there is an accident, the pilot(s) are held responsible.
PILOT ERROR!
Dick
WALTER BREWER
--
Brad Templeton
done by a system a bit more complex than the ordinary level flight
autopilot.) It's rare in part because if conditions are unusual, the
crew will land manually. Just touching the yoke will disengage the
autopilot. The throttle control I think takes a more complex
adjustment to turn off. There have also been incidents where
autopilots were disengaged, and it was judged that keeping the
automatic system on would have done better. Eastern 401 crashed
because the pilots did not realize they had turned off their
autopilot.
There have been some non-crash incidents on Quantas with autopilot
failure causing strange activities. The autopilots were
disengaged. I believe in the recent Buffalo crash, the pilots left
the autopilot on in severe icing when it is not designed for that, and
this contributed. There are other incidents of pilots trusting
autopilots where they should not. (The ones who overshot the airport
150 miles on NorthWest were quite probably asleep, but won't admit
it.)
Of course often ILS landings are done in zero visibility so the pilots
can't see anything (except their instruments) if they take over the
aircraft. They are still landing based on what their computers and
instruments tell them.
The USAir landing in the Hudson river was on "manual" but it was a fly-
by-wire aircraft that had specific software modes to do the right
things in a water landing, and it has been advanced that a regular
human pilot probably would not land the plane safely on the water
without the software systems doing their thing. I believe there had
never been a no-injury water landing of a commercial jet prior to this
accident, and of course the aircraft was still destroyed but all
aboard got out -- because they were surrounded in minutes by rescue
boats.
Richard Gronning
Approaches? Another matter! Remember the icing accident in Buffalo where the plane was on autopilot in icing conditions.
It's a good question Walt, because people should know that there can be a problem with a, "Change of command." What happens if a vehicle is traveling along a highway and the "driver" decides to disconnect the auto-controller? In 99% of the cases, nothing unusual. But, how about unusual traffic conditions? Icy roads? Fog? What happens if the controller wants to go in one direction and the live person wants to go in another?
Dick
On 10/11/2010 4:23 PM, WALTER BREWER wrote:
Has there been a commercial aircraft landing accident when the autopilot was in control?Walt Brewer
----- Original Message -----From: Richard GronningSent: Monday, October 11, 2010 4:15 PMSubject: Re: [t-i] NYTimes.com: Guided by Computers and Sensors, a Smooth Ride at 60 Miles Per HourAll modern commercial aircraft can land themselves.
If there is an accident, the pilot(s) are held responsible.
PILOT ERROR!
DickFor more options, visit this group at http://groups.google.com/group/transport-innovators?hl=en.
Richard Gronning
but none that I know during an auto-land. I wonder which aircraft
disconnect the autopilot with a touch...(???) Boeing aircraft autopilots
can be over-ridden with a great deal of effort, but stay engaged. It
seems to me that any autopilot that would disconnect at a touch would
pose an unsafe condition, (Unless you mean that the touch is to the
disconnect button.) Auto-throttles disconnect when you touch the
disconnect throttle button. It isn't very complicated. Thumb-operated!
Dick
Michael Weidler
| Perhaps this is why they figure PRT ought to be pedal powered and only worth $1M. --- On Mon, 10/11/10, Brad Templeton <bra...@gmail.com> wrote: |
|
|
Jerry Schneider
robocars, I come up with the same question that is often raised about
the feasibility of some dualmode concepts.
If the capacities of these facilities is fully utilized, then will it
be necessary to greatly increase the number and capacity of the
exits, entrances and parking facilities accordingly?
My impression is that the freeway system in most large urban areas is
now clogged in various places at various times of the day, mostly due
to capacity limitations (bottlenecks), high density activity centers
- with random locations clogging up due to accidents, bad weather
conditions, etc. The same is true of some major arterials. One can
see such results, in real time, as on the current travel speed maps
now available in many urban regions.
Suppose you double or triple the volumes with robocars on these
facilities? What are the implications of doing so for access, egress
and parking capacities?
Some of the clogging is due to high trip volumes heading for or
leaving from high density major activity centers at certain times of
the day. Since rapid changes in land use are not likely, would this
kind of congestion become even worse? For example, some have
suggested that instead of building another tunnel under the Hudson
River, you should just build more office and other buildings in
northern New Jersey and use Cisco's Telepresence technology as a
substitute for personal meetings in Manhattan.
If this problem is a "fatal flaw" in the dualmode realm, would it not
also apply to a vast increase in robocar travel volumes or is there
sufficient "unused" capacity in the system (including access, egress
and parking) to make this a non-problem?
- Jerry Schneider -
Innovative Transportation Technologies
http://faculty.washington.edu/jbs/itrans
Jack Slade
Even now, when an accident happens on the street that an exit lane takes you onto, there is nowhere to go. You have to stop, and all of the traffic behind you grinds to a halt also, sometimes for miles. Double the # of vehicles per lane just doubles the length of the backup.
I don't think most of the people working on these projects ever think ot through to it's final stage. There are so many problems with at-grade travel that are not associated with driver ability and reaction times that automated vehicle control can only cure a few of them. Sometimes things happen that no driver (or system) can compensate for, and the vehicle goes completely out of control. It happens with aircraft also, except there is no other veichle a few feet behind.
Jack Slade
--- On Tue, 10/12/10, Jerry Schneider <j...@peak.org> wrote:
> -- You received this message because you are subscribed to
> the Google Groups "transport-innovators" group.
> To post to this group, send email to transport-...@googlegroups.com.
> To unsubscribe from this group, send email to
> transport-innova...@googlegroups.com.
Brad Templeton
park, and when they stop they are technically "standing" not parking.
They can stop in front of driveways, fire hydrants and triple park on
the sides of roads.
When road demand is heavy, the cars are on the roads and not stopped,
so all lanes are clear for traffic. When road demand is light, you
need fewer lanes and you can turn the sides of the roads into parking
(standing) spots . I have algorithms that show how all this works on
my web site.
New ITS and robocar tech can do a lot to increase road capacity, even
without the shorter headways that everybody thinks of first. But one
technology that will make a big difference is metering (or both kinds
of cars) so that you don't put more cars on a road than it can
handle. Computerization can make that metering be done at home, not
in the on-ramp.
With metering, not only do you avoid congestion but you avoid
overbuilding of housing. Traffic load gets automatically distributed
over the road network with metering. Combine it with dynamic reversal
of streets and narrow cars and you have a lot of capacity.
eph
streets and if your vehicle fits, you can go. Also, motorcycles and
scooters (narrow cars) are heavily used and fill up all the gaps in
traffic. It looks like chaos, but seems to work. This both means
that driver-less cars can help in some situations, and also means that
they won't help in some heavily traffic-congested places like Paris.
F.
Jerry Schneider
>Robocars (unlike dual mode cars which are human driven) do not need to
>park, and when they stop they are technically "standing" not parking.
>They can stop in front of driveways, fire hydrants and triple park on
>the sides of roads.
Some dualmode vehicles would be automated on an exclusive guideway, and
only be driven for access, egress and parking on non-exclusive
streets. Examples
are Qwiklane and RUF. What do the occupants do when the robocar must park
in a location that cannot be used for several hours?
>When road demand is heavy, the cars are on the roads and not stopped,
>so all lanes are clear for traffic. When road demand is light, you
>need fewer lanes and you can turn the sides of the roads into parking
>(standing) spots . I have algorithms that show how all this works on
>my web site.
What do the occupants do when the robocar parks on the side of a road?
Getting on and off at desired locations and leaving your own robocar just
anywhere are problems, even if the all lanes are clear.
>New ITS and robocar tech can do a lot to increase road capacity, even
>without the shorter headways that everybody thinks of first. But one
>technology that will make a big difference is metering (or both kinds
>of cars) so that you don't put more cars on a road than it can
>handle. Computerization can make that metering be done at home, not
>in the on-ramp.
Currently, people tend to operate according to schedules that are often not
under their total control (e.g. you will be at your desk at 8 am).
Metering would
work only if the whole system's capacity works, including access,
egress and parking
>With metering, not only do you avoid congestion but you avoid
>overbuilding of housing. Traffic load gets automatically distributed
>over the road network with metering. Combine it with dynamic reversal
>of streets and narrow cars and you have a lot of capacity.
Yes, so long as the indirect route you're assigned to will allow you
to get to your
destination at the desired time. I don't find dynamic reversal and
narrow cars very
appealing unless you just go to sleep and wake up in a "standing"
vehicle somewhere
near your desired destination.
It appears to me that Google's robocar concept needs some "systems
thinking" which
includes all components of the physical infrastructure as well as
some reasonable
forecasts of future land use changes. I'd much prefer to use Cisco's
Telepresence
technology for personal meetings that endure the high volume,
jam-packed traffic and total loss of
control of your vehicle that you envision.
Jerry Schneider
>Paris has all that already. Seems like there are no lanes on Paris
>streets and if your vehicle fits, you can go. Also, motorcycles and
>scooters (narrow cars) are heavily used and fill up all the gaps in
>traffic. It looks like chaos, but seems to work. This both means
>that driver-less cars can help in some situations, and also means that
>they won't help in some heavily traffic-congested places like Paris.
I remember very well the traffic chaos in Paris - certainly as scary
as Tokyo. If you mean it "works" because the traffic generally is in motion
OK, but unless you really enjoy chaos, I wouldn't recommend it. I can't
imagine that older drivers could participate without risking their own lives
and those of others. I suspect it is much worse now than in the 90's and
why would it get any better in the future?
Jay Andress
--
Richard Gronning
Of course the REAL distribution is that;
- The pilot(s) get the blame,
- The airline gets the (bad) reputation,
- The insurance companies pay the claims.
Dennis Manning
Jerry Schneider
>Jay and Dick:
>
>One part about Dick writing about autopilots on aircraft was the
>problem of switching from autopilot to partial autopilot to pilot
>control and vice versa. I should think this would pose problems in
>robocars. I can imagine back seat driver types sitting there nervous
>as a cat second guessing what the robocar is doing and interfering
>inappropriately.
>
>Just because a pilot is held responsible does it translate into who pays?
I would guess that it depends on the fine print in the insurance
policy - certainly the manufacturer might be at fault, as well as the
pilot, airline, air traffic control and the weather and all sorts of
other "causes". I think the "can the passenger override the robot" is
a very important point. Under what conditions would it be enabled?
What would the fine print say about it?
Brad Templeton
a building or inside the building, and then go off and do other
things, including standing and waiting somewhere, renting itself out,
charging itself and so on. It may stand and wait near where you are
if space is available, or further away if not, coming closer on
command or when your approximately scheduled departure time draws
near. Spaces thus don't need to be that near to the people who will
be using the cars (if they own a car, that is.) As such space usage
can be distributed.
In parking lots, robocars can park at valet density. If they are
single person cars, you can probably put 5 or 6 of them in the square
footage used by today's cars in non-valet lots. But you can also put
them on the streets as I described, as the lanes are not needed at low-
traffic hours.
Dual mode vehicles, if they are not automated, must be parked near
your destination and stay there. They would need automation to go
back to the track.
On Oct 12, 12:14 pm, Jerry Schneider <j...@peak.org> wrote:
> At 09:55 AM 10/12/2010, you wrote:
>
> >Robocars (unlike dual mode cars which are human driven) do not need to
> >park, and when they stop they are technically "standing" not parking.
> >They can stop in front of driveways, fire hydrants and triple park on
> >the sides of roads.
>
> Some dualmode vehicles would be automated on an exclusive guideway, and
> only be driven for access, egress and parking on non-exclusive
> streets. Examples
> are Qwiklane and RUF. What do the occupants do when the robocar must park
> in a location that cannot be used for several hours?
>
> >When road demand is heavy, the cars are on the roads and not stopped,
> >so all lanes are clear for traffic. When road demand is light, you
> >need fewer lanes and you can turn the sides of the roads into parking
> >(standing) spots . I have algorithms that show how all this works on
> >my web site.
>
> What do the occupants do when the robocar parks on the side of a road?
> Getting on and off at desired locations and leaving your own robocar just
> anywhere are problems, even if the all lanes are clear.
parking/standing, the occupants left back at their destination, and
the car will return to their point of exit (or as a small electric, go
inside the building they are in if it has a path for this.)
You don't leave the car somewhere, it takes itself to where it will
sit if you are not renting it out. It can sit in any ordinary parking
spot of course, but also in many other spots, including the entrances
of driveways and in front of hydrants. It would leave immediately if
somebody needed to use the driveway.
>
> >New ITS and robocar tech can do a lot to increase road capacity, even
> >without the shorter headways that everybody thinks of first. But one
> >technology that will make a big difference is metering (or both kinds
> >of cars) so that you don't put more cars on a road than it can
> >handle. Computerization can make that metering be done at home, not
> >in the on-ramp.
>
> Currently, people tend to operate according to schedules that are often not
> under their total control (e.g. you will be at your desk at 8 am).
> Metering would
> work only if the whole system's capacity works, including access,
> egress and parking
not allow more people on the highway than it can handle, and they wait
in on-ramps and city streets. The digital alternative is also to not
let more people on the road than it can handle, but to open up other
places for them to wait, including at home. If they need to get to
work at 8, they had better be ready to leave in time to get there,
including any metering waits. It's possible to give people their
metering wait in advance with certain types of systems (bidding,
lottery.) That would be handy, so you would know when to leave.
>
> >With metering, not only do you avoid congestion but you avoid
> >overbuilding of housing. Traffic load gets automatically distributed
> >over the road network with metering. Combine it with dynamic reversal
> >of streets and narrow cars and you have a lot of capacity.
>
> Yes, so long as the indirect route you're assigned to will allow you
> to get to your
> destination at the desired time. I don't find dynamic reversal and
> narrow cars very
> appealing unless you just go to sleep and wake up in a "standing"
> vehicle somewhere
> near your desired destination.
you get to work faster!
>
> It appears to me that Google's robocar concept needs some "systems
> thinking" which
> includes all components of the physical infrastructure as well as
> some reasonable
> forecasts of future land use changes. I'd much prefer to use Cisco's
> Telepresence
> technology for personal meetings that endure the high volume,
> jam-packed traffic and total loss of
> control of your vehicle that you envision.
>
Richard Gronning
the time. Then the airline sometimes gets involved. Most often ALPA, the
union, makes a more thorough study and the FAA ruling gets overturned.
The original question was about autopilot landings, because that
situation is most similar to a driving situation. To begin with, we
already have an autopilot of sorts. It's the cruise control. Certain
cars, like Volvo, will maintain a speed until the car approaches another
vehicle. Then it will maintain a safe distance from that vehicle.. The
cruise control will disconnect by button or by a tap on the brake or
clutch. The driver can override the cruise control with the accelerator,
but it won't disconnect. The overriding of an autopilot on the yoke of
an aircraft takes about 40 lbs of force, as I remember. It can be done,
but not too easily. I also recall that it is easier, once overridden.
Adrenalin plays a part in the ability to override an autopilot. (A check
ride in a simulator is all about adrenalin.)
A simple click of a button is the normal way to override the autopilot
and/or auto-throttles. Each has its own button. It's a good question
about the fine print of insurance companies. I'd bet that it will say
that the operator is responsible. The Google system might have the
ability to navigate in weather - fog. (It's California!) What happens if
it malfunctions? It could be a situation way beyond the ability of the
driver to handle. He/she may not even be able to see the road, let alone
other vehicles.
Dick
Jerry Schneider
>Jerry, the robocar I describe is one that can drop you at the door of
>a building or inside the building, and then go off and do other
>things, including standing and waiting somewhere, renting itself out,
>charging itself and so on. It may stand and wait near where you are
>if space is available, or further away if not, coming closer on
>command or when your approximately scheduled departure time draws
>near. Spaces thus don't need to be that near to the people who will
>be using the cars (if they own a car, that is.) As such space usage
>can be distributed.
OK, I was not considering such possibilities. Depending on the supply
of "appropriate" parking spaces, one might see a lot of robocars driving
around looking for a place to park or stand.
>In parking lots, robocars can park at valet density. If they are
>single person cars, you can probably put 5 or 6 of them in the square
>footage used by today's cars in non-valet lots. But you can also put
>them on the streets as I described, as the lanes are not needed at low-
>traffic hours.
OK, some possibilities here.
>Dual mode vehicles, if they are not automated, must be parked near
>your destination and stay there. They would need automation to go
>back to the track.
Not sure I get your last point. Why would they not be driven back to an access
ramp to enter the exclusive guideway used for automated travel?
Kirston Henderson
>From: Jerry Schneider <j...@peak.org>
>To: transport-...@googlegroups.com
>Subject: Re: [t-i] Re: Google Robocar Announcement
>Date: 13, Oct, 2010, 11:05 AM
>
>>Dual mode vehicles, if they are not automated, must be parked near
>>your destination and stay there. They would need automation to go
>>back to the track.
>
> Not sure I get your last point. Why would they not be driven back to an access
> ramp to enter the exclusive guideway used for automated travel?
>
drive now and they park in the same parking spaces as the current cars. If
we use our CarFerries , the dualmode capability is realized using the same
cars we now drive.
Personally, I continue to see the idea of robocars of some pepple at
Google and the U.S. DoT as unrealistic pipe dreams that are unlikely to see
practical use. There are simply too many unpredictable elements such as
weather, human operators and wild animals that can not be controlled or
anticipated. Another major hurdle is that of providing the essential
redundancy of sensors, processing, and signal paths plus fail, fail-safe
voting logic to allow them to work. This task is difficult enough in
constrained guideway-based vehicles. You also need to consider the added
cost to the vehicle to include all of this.
I spent a significant amount of time during my last years at Lockheed
Martin working on electronic systems design for Unmanned, Autonomous Combat
Aircraft that provided me with a great deal of direct experience with such
matters and there, we only had to worry about possible loss of the air
vehicle and not human passengers or people on the ground. On top of that
the government customer could afford to pay for all of the needed
redundancy. That cost situation is a little different from the typical
automobile customer budget.
I don't believe that people at Google or the the U.S. DoT share that
experience.
Kirston Henderson
MegaRail® Transportation Systems
WALTER BREWER
Have there been autopilot in control landings where an autopilot failure
occured----and either resulted in manual take over or accident?
The Feb 2009 "Continental Airline" accident near Buffalo, and 50 deaths
interacts with a lot of this.
From near the beginning, as noted, pilot error was blamed. And that was the
final verdict, though the reasons morphed into insufficient training, or
min. hours experience for local airlines, etc. The pilot had about 1,500
hours, and co-pilot about 500.
I got interested as a way back before autopilot days because it was hard to
believe even a 1,500 hour "inexperienced" pilot would react incorrectly,
pull back the yoke to raise the nose after being warned by the stall warning
system and the "stick pusher" operating to push the nose down. Was there
some other malfunction that resulted in even slower airspeed, stall a roll
over 90deg, into a spin of about one turn before hitting the ground?
Apparaently not. Neither pilot had stall tranining with this aircraft. He,
she, may have interpreted the pusher to nose down was a malfunction he
should correct.
I looked at all the "black box" recording. (I wonder who maintains all 40 or
so).
Indeed the autopilot has on against rules during what was called moderate
icing. It disconnected automatically at stall warning. There had been an ice
warning a few seconds earlier, and I don't know how such works.
But it raises the interesting point you may just have answered; with
autopilot on, why did speed reduce to unsafe in the first plane. I gues all
autopilots do not control speed as well as aircraft attitude etc etc. That
subjest did not come up in the analysis I saw.
Icing was not blamed. Though I suspect it might have slowed the aircraft
down more than usual. The two pilots were chatting befor thes all started,
and their inattention is blamed. The co-pilot remarked this was her first
experience flyng in ice.
That's backgrowd; now to blame, insurance, etc.
Colgen the airline was under contract to Continental. So tight, Continaental
apparently never entered the blame scene.
As fas as I know insurance paid vicims relatives, and the pilot's estates
never entered the picture. I suppose pilot's contracts cover that point.
Relative of the deceased were mostly from the Buffalo area. They formed a
very tight bond, and as an active group for more than a year lobbied FAA etc
to do something about pilot experience, training, etc. With help of
Congressional reps they were successful in changing those rules, despite the
finacial impact to the airlines. (Though, todays paper has a story that
seems to say a committe set up to implement the new rules, is headed by a
rep from one of the airlines, and are moving the experience level back
again!)
Long story, but this seems to be the case of misuse by a hman who
intervenied with a feature of the more or less automated system to create
the problem.
Walt Brewer
Lo
----- Original Message -----
From: "Richard Gronning" <rgro...@gofast.am>
To: <transport-...@googlegroups.com>
Sent: Wednesday, October 13, 2010 10:06 AM
Subject: Re: [t-i] NYTimes.com: Guided by Computers and Sensors, a Smooth
Ride at 60 Miles Per Hour
Marsden Burger
Best wishes,
Marsden
-----Original Message-----
From: transport-...@googlegroups.com
[mailto:transport-...@googlegroups.com] On Behalf Of Jerry Schneider
Sent: Wednesday, October 13, 2010 12:05 PM
To: transport-...@googlegroups.com
Subject: Re: [t-i] Re: Google Robocar Announcement
OK, some possibilities here.
Jack Slade
Not much. We seldom get snow higher than a single-storey building.
|
|
|
|
eph
possible? How long will it take to certify the software? Will a
company be so bold as to deploy fully driver-less cars before
penetration of all the other related gadgets have been fully
monetized? We still don't have plug-in hybrids available, but they've
been technically possible for many years. If certification is not
needed, will the company fold because of an unanticipated software
glitch even if there is driver override?
Apple Founder: Toyota Problem is Software
Wozniak Says Problem with His Runaway Prius is Not the Gas Pedal
"The day after Toyota insisted problems with random acceleration in
its cars had been fixed by a recall that will replace gas-pedal
assemblies, Apple computers co-founder Steve Wozniak told ABC's Brian
Ross that he believes the problem with his Toyota was not gas pedals,
but software. Wozniak also said he been trying to get the attention of
Toyota and the government's National Highway Traffic Safety
Administration for several months without success. "
http://abcnews.go.com/Blotter/RunawayToyotas/apple-founder-toyota-problem-software/story?id=9728007
F.
On Oct 13, 1:23 pm, Jack Slade <skytrek_...@rogers.com> wrote:
> Not much. We seldom get snow higher than a single-storey building.
>
> Jack Slade
>
> For more options, visit this group athttp://groups.google.com/group/transport-innovators?hl=en.
> --
> You received this message because you are subscribed to the Google Groups "transport-innovators" group.
> To post to this group, send email to transport-...@googlegroups.com.
WALTER BREWER
Walt Brewer
----- Original Message -----
Brad Templeton
On Oct 13, 9:05 am, Jerry Schneider <j...@peak.org> wrote:
>
> OK, I was not considering such possibilities. Depending on the supply
> of "appropriate" parking spaces, one might see a lot of robocars driving
> around looking for a place to park or stand.
parking compared to today. Only cities with very little parking like
Manhattan will have an issue. I actually think we'll be taking the
old parking lots and turning them into parks and buildings.
That's due to many factors. Today there are 3 parking spaces for
every car out there. In the robocar world you can almost get away
with 1, or if there are a lot of single person valet-parking robocars,
0.4 of today's spaces. But that's if the number of cars stays the
same. Robotaxi service, car share and people hiring out their own
cars when they are not using them could reduce the number of owned
cars dramatically. (The reason NYC has so little parking is that
most people do not own cars, and they share, ie. use taxis.)
There are also plans for vehicles like the Twill or MIT City Car which
stand vertically to park, cutting space another factor of 2.
And then there is the surplus of street parking. Every street space
today, plus the entrance to every driveway and every hydrant and every
half-width space that nobody but a SMART can use today. Plus double
that with double parking at off-peak hours, or triple it with triple
parking.
Need more? People can rent out their driveways (not just the
entrance.) You send a signal when ready to leave, or when coming
home, and in 15 seconds the robots have cleared your driveway to find
somewhere else (or wait for you to go through) and you drive away.
But there is so much surplus of parking that I doubt we will do much
of that except in dense CBD areas.
>
> >In parking lots, robocars can park at valet density. If they are
> >single person cars, you can probably put 5 or 6 of them in the square
> >footage used by today's cars in non-valet lots. But you can also put
> >them on the streets as I described, as the lanes are not needed at low-
> >traffic hours.
>
> OK, some possibilities here.
>
> >Dual mode vehicles, if they are not automated, must be parked near
> >your destination and stay there. They would need automation to go
> >back to the track.
>
> Not sure I get your last point. Why would they not be driven back to an access
> ramp to enter the exclusive guideway used for automated travel?
>
vehicle usually needs to stay near where you drive in it, so you can
drive it back to the guideways yourself when you next travel. Unless
you only got it one way, in which case it only needs parking until its
next user comes along. You then have to find another one to get back
to the guideway. In dense areas you can probably do this.