How about something new? I have just one of the possible suggestions

A

Thread Starter

Anand Iyer

After being Nerds in Automation for such a long time now, it seems that something constructive and yet out of this world needs to be done.

So while driving from Khobar to Abqaiq at 6:00 AM, I got this idea....

A wonderful game for nerds....

***************************************************************************
Preface:

In the far far land of autonerdistan, the young princess Fena is partying with her friends. They leave quite late and they are not quite in their senses. They are driving behind a bus and suddenly the front tyre of the bus goes bust, the bus shoots up and hits a overhead bridge and the force of the impact bends the bus into a V shape. The royal car's driver who also secretly partiesd with other drivers, is unable to control the car properly and fragments of the bridge break both the headlights taking out visibility from the driver for some precious moments. In the darkness, the car crashes into the side of the bridge at a very high speed. Princess Fena and her friends are killed in this horrible car crash. The whole land of autonerdistan mourns her untimely death.

The queen orders her son, heir to the throne to build a traffic system which would be automatic, fast and safe and deaths, like that of her daughter's, never occur.

The prince in turn seeks the help of the council of ministers to come up with suggestions within a week. A committee is formed. The Supreme committee for automatic traffic in autonerdistan (SCATA).

The commitee immediately recommends that the traffic system be modified in the following way:

The first part is data acquisition and data provision.

The roads should be made intelligent.

Every traffic sign board should transmit electronically, its meaning to the cars on its roads. That is if there is a 'right turn ahead' signboard, then the same is transmitted electronically to the car. The traffic sign board also further transmits whats the next board and the distance of the next board. The traffic sign board also signals its position in a x,y coordinate system to the cars travelling on it. Each traffic sign board also has a second transmitter to give dynamic information as required by the management systems in second part to all cars. Each traffic sign board has a reciever to recieve information from cars.

The cars in turn are intelligent and give the following signals to each other. Their make, model no, number plate, speed, position (accurate upto lane level). The cars also are to give status signals like they are reducing speed, accelerating and so on. The cars are automatic in the sense that speed and directional movements are automatically controlled. Manual mode is possible only in worst cases (that is human driving).

Intersections are to be replaced with round abouts which cars changing lanes from 4 to 12 points.

The second part is utilizing this information and managing traffic. The comittee recommends a system of collecting all this information and
using a system to: Manage lane changing. vehicl and communication systems Break down management. VIP services and traffic management for VIPs.
Electronic barricading of roads.

Speed limits are to be made dynamic. If the loads are low, then speed limits are raised and speed limits can also be lowered. No car can go faster than the speed limit prescribed. Total darkness (no headlights) or Zero lux driving systems for war times.

Committe recommends making x,y coordinate road maps and using them in the system for various purposes.

Knowing that this is a new beginning, the SCATA requires the help of automation engineers and their knowledge and practical experience being so important, they announce a competition. Design a system using field products (you can also use products that you specially make, provided you give the design and it is not copyrighted) and systems (the choice of operating system and programming languages is not controlled).

The cirteria for winning the competition are:

zero accidents in all conditions.
average speed of transportation.

You can use any city map.
You may not follow any or all of the SCATA recommendations but make your own rules, own designs and own policies!


***************************************************************************

The competition could be categorised into

best over all design.
best over all design from students (undergraduates).

Best design using product of specific company. (In case some companies are interested in sponsoring some awards). Design period one year (Which is a small amount of time considering most of us have the great advantage of working on weekends too!!!). The whole year the list clarifies and debates on various points of the design..... Members of the list could also contribute minor amounts towards the prize money and the final amount could be quite big!!!

Regards,

Anand
 
Anand:
> A wonderful game for nerds....
...
> The Supreme committee for automatic traffic in autonerdistan (SCATA).

Problem number one in your system: pedestrians, wandering stock, fallen trees. Some things just have to be seen, and vision technology is not up
to the task.

Problem number two: legacy cars. The rest is easy, or at least doable.

> Every traffic sign board should transmit electronically, its meaning
> to the cars on its roads. That is if there is a 'right turn ahead'
> signboard, then the same is transmitted electronically to the car.

This is superfluous. For one thing, video processing is up to this task. For another thing, the cars can have all the signs marked on their maps.

> Manual mode is possible only in worst cases (that is human driving).

Hmm, if you allow manual mode, you lose a lot of the benefits of every single car being automatic. Having to contend with human drivers in the
system will make it much more complex. Especially since a few years after the introduction of the system everyone will be out of practice and the only people using manual mode will be drunk teenagers.

> Intersections are to be replaced with round abouts which cars changing
> lanes from 4 to 12 points.

No real benefit, extremely large cost.

> Total darkness (no headlights) or Zero lux driving systems for war
> times.

Now, there's a system to appeal to a government :)

> Committe recommends making x,y coordinate road maps and using them in
> the system for various purposes.

Obviously, it's not that simple, but such systems already exist.

> The cirteria for winning the competition are:

> zero accidents in all conditions.
> average speed of transportation.

I'm leery of the number zero; it has an unpleasant tendency to translate into an infinity elsewhere in the system.

Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
A
Hi Jiri,

The idea is to draw up a working system.

None of the systems can be 100% safe, but we can almost make up the mark. Of course we can use radar, or IR sensors to detect objects on the road. But in practicality, this may shoot up the cost of the car! Or simply, barricade roads at most points (that does not work even today, people jump over barricades.). And even if you detect you may not be in a position to take evasive action.

However, we can definitely try to solve say 95% of the problems. AND Just design and have some fun, have some competitions and so on.

And the great thing about such an exercise will be that people will come up with so many different ways to solve the problem... thats just wonderful.

But of course, judging from the responses so far, people are too busy for such games.

Lets see if people are interested.

Anand
 
Anand Iyer:
> The idea is to draw up a working system.

> None of the systems can be 100% safe, but we can almost make up the
> mark. Of course we can use radar, or IR sensors to detect objects on
> the road. But in practicality, this may shoot up the cost of the car!
> Or simply, barricade roads at most points (that does not work even
> today, people jump over barricades.).

Yeah, that reduces the problem to something equivalent to a railway, which is easy, and therefore boring.

> And even if you detect you may not be in a position to take evasive
> action.

Didn't the original specification mention a collapsed bridge gumming up the works?

> However, we can definitely try to solve say 95% of the problems. AND
> Just design and have some fun, have some competitions and so on.

95% is useless (and, quite likely, existing tech).

> And the great thing about such an exercise will be that people will
> come up with so many different ways to solve the problem... thats just
> wonderful.

> But of course, judging from the responses so far, people are too busy
> for such games.

I think the people who are doing autonomous cars aren't on automation-l, and as for different ways of solving the problem, I'd recommend a good
AI textbook as a starting point.

Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
Top