After seemingly endless hurdles, each to be jumped by tips from a different web post, my ROS2 ROSbot Dave is sporting a LIDAR that is publishing /scan topics!
One thing surprised me when looking at the scan. In the scan off the bot’s right side (directly up in the screen shot) the scan shows a short section of “wall” with gaps on either side of the short section.
That corresponds to a reflective black metal filing cabinet. Apparently, only beams hitting the metal nearly perpendicular bounce back with enough intensity for the LIDAR to see the obstacle.
I have the feeling these wall “dropouts” are going to make wall following more challenging than expected.
Makes sense. I don’t have anything really shiny like that, so I haven’t seen that. The weekend I actually rebuilt a plywood “roboworld” so I can tweak things in a very controlled environment to start with.
IMHO, for what little it’s worth, that’s an excellent idea.
Can you create a simple(er) environment to get things standardized. Once you’ve got the basics knocked, you can then introduce Dave to a more complex world.
It’s like not re-writing code - you want to reduce the number of “first-order forces” to the absolute minimum to reduce the number of articulating points and variables to a minimum as well.
(CyclicalObssessive hangs head in
shame awe.) At HP they wouldn’t issue screwdrivers to the software engineers so we wouldn’t hurt ourselves.
Dave will have to be content with being a simple mind in a very complex world, and with me to rescue him when he gets in trouble.
Nothing fancy to be sure. First attempt at navigation was an immediate failure - I think my costmap settings extends the buffer zones too far into the space, so there’s no room to move. Project for Saturday.
Dave should hold his
head body head high - he’s making progress very fast in a way more complex environment. I just need to keep it simple.
Charlie stands in awe of your accomplishments too.
When I worked as a QA engineer at Sensormatic, testing access control systems, there were certain divisions:
Software engineers were given pre-configured hardware and were forbidden to touch at risk of immediate dismissal.
Hardware engineers were given pre-configured software with the same restrictions.
Hardware and Software engineers who were able to demonstrate the requisite skill were allowed to work on both and were the second tier - qualification engineering - and received higher pay.
QA engineers had to demonstrate skill in hardware, software, and computer configuration and repair. (PC or VAX)
We also had to prove we would not electrocute ourselves when working with mains power.
We were paid about half-way between the two engineering tiers.
It’s a pity that you won’t try at least some simple construction. You won’t learn if you don’t try.
I compel my granddaughters to have at least a minimal competence with hand and power tools and basic knowledge of various kinds of repairs so that they get taken advantage of by unscrupulous repair people.
Good skill to have for oneself. And for the company it avoids a lot of paperwork
@jimrh Over the years I have tried simple and not-so projects, and have learned to approximate how much time and materials will be required, when I will be happy with my results, when to hire an expert, and how to find a competent expert at a price I am willing to pay.
I’ve even designed and personally built an active-solar ranch home from digging the septic tank, the entire plumbing and electrical, to siding and roofing the project. Wife left me for another woman and stole my horse after that project. I know a few things I’d rather not know about construction.
In the words of a great song - “I wish I didn’t know now what I didn’t know then”