Observations, Insights and Conclusions gained from attempting to create an autodocking process for the ArloBot

[thomasco...@gmail.com]

Hi All,

Observations, Insights and Conclusions that have been gained from attempting to create, through hardware and software, an autodocking process for the ArloBot.

Observations/Insights:

1) ArloBot has been found to be a difficult robotic platform to work with due to its low resolution wheel encoders.
2) The DHB-10 Arlo motor controller is overly sensitive to minor overcurrent conditions that can cause the need for constant resets as the battery voltage begins to drop with use.
3) ArloBot has been found to perform the autodocking process the best at `linear_x `speeds of at least 0.05 meters/second or faster.
4) Unlike the PING Sensors, the present auto-docking hardware/software configuration is not capable of stopping ArloBot relatively instantaneously and takes approximately 1 - 2 seconds to get ArloBot stopped once it has contacted the pogo pin charging array.
5) The faster ArloBot approaches the docking station the better the alignment of the ArloBot docking probe to the docking station pogo pin charging array.
6) Due to the time delay getting ArloBot stopped, once it has contacted the pogo pin charging array, limits the speed at which ArloBot can approach the docking station to 0.04 meters/second or less. The limited ArloBot docking speed greatly affects the alignment of ArloBot to the pogo pin charging array as ArloBot approaches the docking station.
7) ArloBot will usually complete a successful docking 90% of the time.

Conclusions:

1) Find a way to stop ArloBot's forward motion once it has contacted the docking station pogo pin charging array that is equal to that of the PING Sensors Escape function when they detect an object in front of or behind ArloBot. This will allow a docking speed of 0.05 meters/second or greater which will greatly enhance the ArloBot chances of a successful alignment to the pogo pin charging array on the docking station.
2) ArloBot definitely requires a more robust motor controller, like the RoboClaw, to improve its motor control performance and possibly wheel encoder discs with more vanes to improve its speed and distance resolution.

Comments?

Regards,
TCIII 

[bebe...@gmail.com]

Really interesting post. I wish I had time to look into this more with my ArloBot, but my robotics tinkering is on hold for the near future.

Item 2 caught my attention. What problems are you having with the DB-10 motor controller?

There is an old thread regarding DHB-10 motor controller problems that comes to mind that may be of interest if you have not already seen it:

https://forums.parallax.com/discussion/167944/arlobot-dhb-10-motor-board-bug-with-encoders#latest

Regards,
Brad

[thomasco...@gmail.com]

@bebe,

Thanks for the link, much appreciated.

Yes, I have seen that DHB-10 thread on the Parallax Forum, however I do not have an encoder issue as was discussed in the thread. Just an over sensitivity to slight overcurrent conditions when the battery has been in use for a long period of time.

Regards,

TCIII