Singularity News

Business As Usual

Posted by Darin Velarde on

Hello, everybody, it's been a while since I last wrote a post.  It's business as usual at Singularity Robots. The Singularity Bot has had some improvements.  I found and fixed the bug in the algorithm and it's showing promise for being able to navigate and "remember" things.  I ended up writing a Qt application that displays all the runtime information on a remote Linux computer.   I decided against using Unreal Engine 4 for that sort of thing (I was already doing that).  It was just too much overhead and takes too long to compile and start up.  The new...

Read more →

Robot progress is slow but steady

Posted by Darin Velarde on

The Singularity Robot is slowly coming along.  The Motor controller is working along with an MQTT client that is using the pub-sub paradigm to get messages about what to do.  It's implemented with a fail-safe system that will stop all motor operation if messages stop coming in for over 2 seconds.  This is to save baseboards and ankles.  This thing weights over 40lbs, so it would hurt, and we don't want that.  Progress on the navigation has been slow due to a bunch of other things going on but the plans have not changed.  I started working on a learning...

Read more →

The latest on the Singularity Robot

Posted by Darin Velarde on

Everything is in the Singularity Robot's case.  Currently we are using the Odroid XU4, the 5-port ethernet switch, and the STM32F4 discovery board with the  STM32F4DIS-BB (the baseboard for the STM32F4 discovery board), and the STM32F4DIS-LCD (the LCD screen that plugs into the baseboard). We are using a custom MQTT port for the communication to the STM32 which is being used for motor control only at the moment. We are also using a Hokuyo urg-04lx for scanning the area around the robot.  There is still much to do but it's coming along nicely. Next steps: Implement the motor control in the...

Read more →

"Embedded" Linux computers (continued again)

Posted by Darin Velarde on

For those of you with the Rock 2, you might want the below pin mappings.  The Radxa folks were kind enough to send me the mappings that I was looking for.  I have not fully tested them all but I have tested the ones that are the same as the original Rock and they work great!  So at this point, it's a shootout between the Odroid and the Rock2.  Let the testing begin.  They are both functioning as they should with no issues whatsoever.  (Note that I have not used I2C or serial yet on either of them) GPIO0_A0 = 0GPIO0_A1...

Read more →

"Embedded" Linux computers (continued)

Posted by Darin Velarde on

We have the Odroid in hand and it's working great.  I was able to get my whole software stack running on it, which is really great. In the mean time I've been conversing with the Radxa folks on the rock2 square and was able to get the GPIO sort of working.  Currently it's using the sysfs system in Linux, which is pretty standard.  The issue now is that I don't have a mapping for GPIO numbers to actual pins.  If anybody has this information, please let me know where to find it as it would be nice to get the rock2...

Read more →