Proximity Sensor selection questions for Propeller Microcontroller

[TheGrue]

I am conceptualizing a design for a new robot based on the Propeller and the Motor Mount & Wheel Kit. The robot will be both an indoor and an outdoor bot that travels between buildings by navigating its way through a parking lot and on sidewalks. I am looking for ideas for sensors to detect objects and obstacles that could trap the Bot. I do not want to overload the bot with sensors but I have specific desires and a specific obstacle course in mind. I am not looking for help on navigation from point A to point B right now.

The bot will have to avoid driving off of raised sidewalks and curbs, follow the edge of a curb to a wheelchair ramp, cross a parking lot that may have cars or other obstacles, follow another curb and up another wheelchair ramp to the next building entrance. I was also thinking of having sensors in front of each wheel to avoid things like potholes or one of the many 90 degree T's in sidewalks that the wheels may drop into and get stuck. Also I thought that they could be used to detect if it is driving into rising water during bad weather. I am not needing water proof sensors, just water resistant and inexpensive if they get ruined by wet weather.

I have heard that PING))) and other ultrasonic sensors are not good to detect cliffs or stairs if pointed at angles like 45 degrees. I do not know the top speed of the Motor Mount Kit that Parallax offers but I would want to stop the bot before it went off a curb. I am intrigued by the LV-MaxSonar-WR1 for being long distance and weather resistant. Also it would be good to take indoor readings if I had one facing all four directions. It could be used to identify corridors and for indoor mapping. The down side to this sensor is that 0-12 inches is read as 12 inches. The up side is that it runs on 3-5V and has several types of outputs that would easily interface with the Propeller. I also see that the Sharp line of IR sensors are popular and I purchased two for testing (GP2D12 and GP2D120). I set them near each other to scope their outputs and found out that they interfered with each other when pointed in the same direction or looking at the same object. The up side is that they are cheap if they get toasted by the weather. The down side is that each runs on 5V and needs A/D conversion to get useful distance readings.

I am done rambling now and wish to hear lots of ideas...LOL

 

[Sceadwian]

I'm a big fan of the multi-core concept on those propeller chips, great for robotics users. I was thinking about buying some but I don't like Parallax and for general micro controller use they don't have much of an advantage over more standard micro controllers.

Onto the post =) One way to add 'free' sensors to your robot is to put current sense resistors in line with every motor so you can sense the amount of current (hence torque) the motor is using. If something is using an unexpected amount of torque it's encountering resistance which you can react to.

 

[TheGrue]

Sceadwian,

You mentioned "for general micro controller use they [Parallax] don't have much of an advantage over more standard micro controllers".

Which micro controllers are more standard?

 

[blueroomelectronics]

PIC and AVR are more standard (popular) these days.

 

[Sceadwian]

8051 as well, still used extensively.

 

[Sceadwian]

Parallax themselves dont make anything that I'm aware of, the propeller chip might be a first, they just rebrand someone elses's chips. Their basic stamp line is based on PIC micro controllers, and the SX key series they put out is a Scenix micro controller. The basic stamp line are crippled pics though because they load tokens from external memory. Slow as molasis. Don't know much about the ones that use the Scenix, but the development software for Paralllax stuff and the prices they charge for the bulk of their products is very high because they're trying to take advantage of a niche hobby market.