rec.autos.simulators

I am looking for any kind of driving (car/truck/scooter) simulator
that I can use in a research project on drowzy/drunk/sober driver analysis.  
I need to be able to control the speed, steering, etc.

Can anyone help?  Thanks !

There are demos of not too picky "simmers" like Microsoft's Midtown madness,
I personally never got any of the Need For Speed (BY Electronic Arts) to
ever run on my computers it seems.  especially the demos...

Live For Speed (LFS) by totally different group, would be more my choice as
it has reasonable Physics, free in evaluation mode (limit on cars & tracks)
it is limited until you pay $ (actually euros) to unlock it, so wouldn't be
too bad, plus is "City" orientated.  other ones might be a bit difficult to
get running, and with you using it funny?  least LFS should run, with a good
wheel joystick?

    Now about controlling, you dont mean not as a driver right?  like
drivers ed teachers have gas, brake, & sometimes a wheel to take over??
that I havent seen anywhere...  cause nobody's life at risk either LOL.

Racer and Netkar would probably be suitable.

Jason

An example of what I want is like this:
http://www-2.cs.cmu.edu/afs/cs/project/space/www/hcs/simulator.html
It is a software driving simulator devised by people in CMU long time ago
for research only. But they cant find the code now.

Any suggestions?
thanks a lot

Haven't tried it myself, but I hear it's pretty nice.

HTH,
Kendt

PS - would be funny to add routines to simulate narcotics, speed, and
LSD ;)!  How about the floating goats from No One Lives Forever ;)?

You can drive many of the PC entertainment software titles that include
detailed telemetry. You could measure lap times or number of accidents
per lap, or etc.

- Matt

Good luck,
Kendt - a biology BA who works with computers, so any research project
gets my interest up.

Any simulators can save these data into files so that I can read?
(I tried LFO and Racer. Seems LFO doesn't provide that kind of data, and
I cannot get Racer work on my pc.)

thanks a lot

Select for coders
now some additional selections should come up
Select RAF FORMAT (for programmers)

Replay Analyser file version 2
this is the file format for the replay analyser. If you are capable of some
programming you can write a specific program for your needs.

All of the programs listed have the ability todo what you are looking for, they
all save the information, they just do not display all the things that you are
looking for... with some creativity on your part you will be able to make any
of them do what you want...

Best of luck, sounds like an interesting project.

Rich L.

Here's a link to a group that is doing the same kind of research:
  http://www.biomedicalinstitute.org/newsarticle&record_id=29

Years ago, Daniel Cox used a modified Atari Hard Drivin' (arcade game)
for this kind of work, now it looks like they are getting a dedicated
simulator.  He published a number of papers...should be in some medical
indices (but I don't have any links).

-- Doug Milliken
   www.millikenresearch.com

thanks a lot

...

The latest Racer alphas (v0.5.2) record the 2 first entries; the
offset wrt the median (lateral position ranging from 0..1) is trivial
to add.
It is specified through logging, which can be toggled on/off with F3.
It records in a very simple format which specifies the fields in the
first line, then the data, like this:

time steer0 steer1 u0 u1
1200 0.5 0.5 0.1 0.2

Where time is in milliseconds, steering is in radians, u would be in a
range of 0 (left side of the road) to 1 (right side).

I've used this for example in a rough telemetry analysis tool to
examine a high-frequency force feedback defect which you could feel in
a high-end wheel, but not consumer wheels. And we use it with ADAMS
for example to analyse car behavior.

I could do the lateral log feature and send you the exe, but that's an
v0.5.2 beta so requires some tweaking to run perhaps, and I see you've
been having trouble (perhaps a firewall issue, check
c:\sims\racer\QLOG.txt for possible hints).

Still, as others indicated, you can also use existing sims. It's
mostly the extent of the telemetry data that is the limit.

Ruud van Gaal
Free car sim: http://www.racer.nl/
Pencil art  : http://www.marketgraph.nl/gallery/

...

Ah btw, do you intend to let actual people drive, or do you want to
simulate a 'drunken AI'?
That last one would be interesting, although there's nothing like a
human to simulate 'drunk controls', if only since it would be more
fun. ;-)

Racer doesn't have real support to input speed/steering externally
other than joystick/keyboard/mouse, but you might be able to fake it
through mouse warping for example to simulate left/right,
throttle/brake. That's how my AI does things a bit anyway (not
warping, but 2 axes, 1 for steering, 1 for 'acceleration').

Ruud van Gaal
Free car sim: http://www.racer.nl/
Pencil art  : http://www.marketgraph.nl/gallery/

Hi Ruud,

* <joke> Variable "u" is already taken for longitudinal velocity--see RCVD
Chapter 4.  You will have to pick another letter for lateral offset<grin>.

* When we did our informal experiment, we actually got drunk (monitored
with a breathalizer) and kept track of scores in Hard Drivin'...  One of my
friends has the data, I think he uses it as an example of "confused data"
for a statistics course he teaches (internal to the computer company where
he works).

* Can you say what kinds of things you do with ADAMS?

-- Doug Milliken
   www.millikenresearch.com

<snipped>

Hehe, what I'd suggest is you send everyone a voucher so they can
bring back RCVD and get a new copy where chapter 4 corrects 'u' into
something other. ;-)

Cool. I think I'd just get sick in our motion platform, but the idea
is interesting (only from a scientific standpoint ofcourse, ahem). :)

Actually, I don't use it myself, but I'm working with a company which
does virtual prototyping and uses that. So they work out forces on
prototypes for cars, suspension analysis and such in software.
They're also working with Racer & FCS-CS here in Amsterdam to create a
decent car that behaves realistically. They have access to things like
MF-Tyre to create Pacejka sets from sample data for example.

The top picture at http://www.f1simulator.com shows the motion
platform plus high-end steering wheel (it can rip your arm off if you
don't limit it) and Racer in action. The belts are really useful. ;-)

Racer exports telemetry data which can be imported in ADAMS, IIRC, for
analysis. The steering wheel is quite more sensitive than anything I
have lying around (a bunch of consumer wheels), so you tend to miss
high-frequency effects (some of those occurred some time ago with
lateral relaxation experiments).

The wheel/platform is controlled through a network and it's quite fast
at 100Mbit - I know you're more into direct hardware control but with
100Mbit ethernet the time to get to the wheel is probably shorter than
the next frame being drawn onscreen. :)

The platform is meant to be the highest end thing available, and it
sure does feel nice I must say.
Sorry to digress, but it's just exciting stuff. :)

Ruud van Gaal
Free car sim: http://www.racer.nl/
Pencil art  : http://www.marketgraph.nl/gallery/