Whither will DB1 wander
It looks like DB1 will start chuntering along shortly. At the outset the goals for this robot were not defined, so I'm are not clear what it will end up as or indeed what will be found along its development path, and that leaves open many possible experimental routes to indulge in ? . Will it include the rotating LIDAR module Bill introduced a year back, will Pixy 2 be used to recognise little green blobs, what will Bill be putting on that turntable?
So I'm going to put in a plug for somethings I would like to see developed, and somethings I would plump for avoiding.
1. the ability to navigate to a charging station to self charge. This should include the prevention of overcharging. The more I think about this the more I see it as an imperative. Will I remember (or be minded) to keep plugging and unplugging a charger once the first fun wears off.
2. navigate by communicating with a local positioning system. (local object avoidance as well of course)
3. move away from Arduino and C and use microcontrollers running micro-python. (did I hear a sharp intake of breath)
1. Using pre-baked ROS. Much better for our understanding and tweaking to to write all the code for ourselves.
2. Don't stray into AI, that appears to be big black hole for computer resources for doubtful usefulness.
My desires for the ongoing experimental paths for DB1 all tosh? doubtless Bill will find all suggestions to be of interest at least for a laugh.
And, when all the experimental fun has been had with DB1 I give a plug for a son of DB1 to be build as robot with a tightly defined purpose. (e.g. pick up mail from the floor and put it on the table, or prod the sleeping man if he has been inactive in his chair for too long). One can learn a lot by seeing how problems are overcome to achieve a goal.
My own robot progress has stalled as other things have priority but I hope to get more stuck in to this hobby by the end of 2020. But I still take some time to follow along and and stash away some ideas for when I can back into electronics.
Hey Bill - go at your own pace and make sure you enjoy 2020, thanks for all your interesting articles, all and every one.
For those like me who haven't known about using Python to program a microprocessor.
The idea of ROS I think is to provide a framework to make use of other people's stable code to save having to write it yourself. For example two useful functions for a robot I wouldn't be able to write code for would be the ability to find and identify faces or translate the spoken word. ROS also provides a common framework to share your code with others. Having said that I will not be using ROS because after spending some time looking at tutorials I have found it too complex compared with the simple code I have written myself.
AI can be simple or as complex as required for the task.
I think the robotic vacuum cleaners with vision and/or LIDAR set the benchmark as to the basic starting point of any domestic robot. It would also be desirable that a robot would be self cleaning and able to monitor its own states for mechanical problems.
Yes this ai thing is surely complicated I am so far behind on the db1 system I'll have to play catch up and since the 3d printer took over the panda I can no longer use ros on windows it looks like I need a SBC that runs Linux if anybody knows of one let me know . in the mean season I just wrote a simple sketch to move the robot through some motions forward backward wax on wax off .??
Yes the Jetson is certainly next in line im trying to keep it under the 200 dollar mark the Jetson is a good option it would be nice if it could accommodate a 7 in screen I don't know if its possible but it would be nice to be able to see some of the sensor data being displayed on the screen as it comes in from the system.?? as it cycles through the different sensors in the array .