Skip to content

Carbot Telemetry Results

April 27, 2015

I ran the Carbot with the XBee reporting the sensor results this morning.

In the video you can see it make one good turn and then one wild turn. The front sensor initiates a turn if the proximity reading gets above 160 and keeps turning until it drops below that. The good and bad turn data are shown below. The @nnn is a time-to-live counter, allowing carbot to run 500 cycles, the are the proximity readings and the resulting direction either Forward or Left.

9:52:58.712> @276<148 F>
9:52:58.774> @275<168 L>
9:52:58.961> @274:<@273:<197 L>
9:52:59.086> @272:<201 L>
9:52:59.149> @271:<176 L>
9:52:59.273> @270:126 F>
9:53:01.145> @253:<159 F>
9:53:01.270> @252:<184 L>
9:53:01.333> @251:<201 L>
9:53:01.457> @250:<228 L>
9:53:01.520> @249:<260 L>
9:53:01.645> @248:<258 L>
9:53:01.707> @247:<217 L>
9:53:01.832> <115 F>
9:53:24.520> @39:<174 L>
9:53:24.644> @38:<186 L>
9:53:24.707> @37:<201 L>
9:53:24.769> @36:<383 L>
9:53:24.894> @35:<193 L>
9:53:24.956> @34:<102 F>

I note that the good turn got five Left turn cycles while the bad turn got six. Other good turns in the data show four cycles. I could also tell that the car had been veering left before the bad turn and veering right before the good one(veering left or right is what the car does to keep its distance from the wall). In the last turn shown in the video – which is a bit too sharp, there are five cycles of left turn and it was veering left before the turn.

I measured the distance from the wall when the front facing infrared sensor was reading 160 and it was right at the 18 inches I was aiming for. Somewhat to my surprise though, when the car was at an angle to the wall it read 160 at a slightly greater distance – I had actually figured it the proximity reading would drop off when the sensor was at an angle but even at quite a sharp oblique(22 degrees) it would read 160 at a distance of 20 inches. The ultrasonic by comparison drops off quickly at anything less than 45 degrees.

You would think that four-six cycles would be granular enough for control but one easy change would be to reduce the length of the turn cycles.

UPDATE: I tried a 30ms turn cycle which generally seemed more successful although my telemetry wasn’t working and the motor batteries ran out. I’ll think about what i can do to improve forward speed and try again tomorrow once the batteries are recharged.

Advertisements

From → Uncategorized

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: