...at least a little bit ;)
Tobi wrote me an email (unfortunately in german) where he cleared up my mind and told me again the strenght of our concept using only the enviroment and not the inside of a car.
The strenght in our sytem is that it owrks by itself. It doesn´t matter, what techniques is inside teh car. And if we would let the crossroad comunicate with the car, then we could also say, that the car not only warns the driver but directly controlls the car. But that is something we don´t want to have. We want an intelligent intersection and we are gonna make it!
So just forget my "some thoughts" and I apologize for confusing you...
cya on monday
Saturday, November 15, 2008
Friday, November 14, 2008
Friday Meeting
The next class meeting will be Sunday in which we should have a nearly finalized idea to share with the class. After considering Andy's comments on our current progress we have decided to add a windshield component to the "intelligent intersection" system that is underway. We now need to refine how these two components work together, suggesting that we work with Flux's group.
Lauren and Jenny will research further blob-detection and the materiality of this project.
Tobi, the new diagram is really insightful and suggested some really great ideas. We are thinking to do away with the zebra stripes as your diagram suggests and reduce the signage to single lines. Also, is it possible to light up the pedestrian as he or she crosses?
We will discuss everything further Monday.
Lauren and Jenny will research further blob-detection and the materiality of this project.
Tobi, the new diagram is really insightful and suggested some really great ideas. We are thinking to do away with the zebra stripes as your diagram suggests and reduce the signage to single lines. Also, is it possible to light up the pedestrian as he or she crosses?
We will discuss everything further Monday.
Some Thoughts
Hi,
this time I can´t come up with a diagram or new picture because I don´t really have a good idea. The last critique on the curb-lights was the problem with the visibility. Trying to find better solutions I always come to the point, that having this kind of warning system inside the car (the position indicator of the pestrians at the sidewalk) and even on the windshield could be working better. There are several reasons for this:
Why not outside:
1. Putting light not on the sidewalk but somewhere else, where it is better to see, could also affect other drivers, who don´t want to turn. They shouldn´t see those lights.
2. Not only parking cars could be in the line of sight but also other driving cars.
3. A lighting system "outside" is good at night, but is difficult at daytimes.
4. The driver already has to pay attention to so many other things in the enviroment (lighst, sogns, cars, etc.) that one thing more is making the situation even more complicated.
Why inside:
1. The driver iss always looking "through" the windshield, so he´s also looking "at" the windshield. Why not putting an importent Information at a spot, where he´s is looking at anyways?
2. Having such a system inside a car can filter the information from outside for each driver speciffically and individually.
3. If you don´t want to make a turn, the system is off and you can focus on other traffic actions.
4. You could even use other signals like sound for example, to indiacate a dangerous situation.
Referring to Tobis 2nd Animation:
- Making the corner lights brighter, when a pedestrian is approaching the crosswalk iss good, but not working at daytime.
- The lights on the sidewalks are ok, but what happens, if you have somebody driving in front of you?
- What happens at the Zebra-stripe, when people are coming from both directions? And do we need the signal there or is it allready to late?
I think the solution for our problem is somewhere between Tobis Animation and an additional warning system inside the car. And I also think, that it is no problem, that Ozs Group is also working on the Windshield: As I understood, they are working on a HUD which tells the driver a lot of general Information via the Windshield.
Please don´t get me wrong: I liked all our ideas till now, but I am just a little bit concerned If we are moving totally the right way.
Just some thoughts...
Greetz andy
this time I can´t come up with a diagram or new picture because I don´t really have a good idea. The last critique on the curb-lights was the problem with the visibility. Trying to find better solutions I always come to the point, that having this kind of warning system inside the car (the position indicator of the pestrians at the sidewalk) and even on the windshield could be working better. There are several reasons for this:
Why not outside:
1. Putting light not on the sidewalk but somewhere else, where it is better to see, could also affect other drivers, who don´t want to turn. They shouldn´t see those lights.
2. Not only parking cars could be in the line of sight but also other driving cars.
3. A lighting system "outside" is good at night, but is difficult at daytimes.
4. The driver already has to pay attention to so many other things in the enviroment (lighst, sogns, cars, etc.) that one thing more is making the situation even more complicated.
Why inside:
1. The driver iss always looking "through" the windshield, so he´s also looking "at" the windshield. Why not putting an importent Information at a spot, where he´s is looking at anyways?
2. Having such a system inside a car can filter the information from outside for each driver speciffically and individually.
3. If you don´t want to make a turn, the system is off and you can focus on other traffic actions.
4. You could even use other signals like sound for example, to indiacate a dangerous situation.
Referring to Tobis 2nd Animation:
- Making the corner lights brighter, when a pedestrian is approaching the crosswalk iss good, but not working at daytime.
- The lights on the sidewalks are ok, but what happens, if you have somebody driving in front of you?
- What happens at the Zebra-stripe, when people are coming from both directions? And do we need the signal there or is it allready to late?
I think the solution for our problem is somewhere between Tobis Animation and an additional warning system inside the car. And I also think, that it is no problem, that Ozs Group is also working on the Windshield: As I understood, they are working on a HUD which tells the driver a lot of general Information via the Windshield.
Please don´t get me wrong: I liked all our ideas till now, but I am just a little bit concerned If we are moving totally the right way.
Just some thoughts...
Greetz andy
Thursday, November 13, 2008
Friday, November 7, 2008
Fridays Meeting (Design and Technology)
We met with two professor's of Design and Technology (DT) and walked in on a demonstration. They have a program that is projected onto a floor pad about 12' x 12' ft2. The floor is equipped with infrared lighting and motion sensing cameras. The minicar senses darkness against the white pad and tracks it shape.
We would like to use a similar technology using Computer Vision/ Blob Detection to recognize shapes in the intersection to trigger the interactive crosswalk.
Hopefully we will be able to animate all this in After Effects...
The DT guys also said the fiber optics embedded with LEDs in asphalt is feasible so we will research how this would work.
Andreas, the only critique we had for the curb and arrows idea is that if a cars are parked next to the curb, it won't be seen. As usual, your presentation was very beautiful and you are very skilled with diagramming. We would like to further develop the idea of bathing the pedestrians aproaching the curb with light, perhaps tracking the pedestrians with a spot-light? or tracking them with intense light, like Jenny S's diagram?
Direction we are heading:
-Interactive crosswalks
-Lighting up the pedestrians
-LED's and fiber optics
-Computer Vision/Blob detection
For next week, lets refine these ideas. We have 4 weeks to go and we are surprisingly right on schedule.
Spreadable OLED's
www.engadget.com
- This could either be spread or "painted" on virtually any surface, leaving an OLED screen that would be just 100nm thick.
- At would also double as a solar panel that'd keep the OLEDs powered, something that's apparently possible due to the "similar, but opposite, principles" that OLEDs and solar cells work on.
- Researchers from Sumitomo Chemical and Mitsubishi Chemical are aiming to have usable prototypes ready within the next two years.
Subscribe to:
Posts (Atom)