Tuesday, April 18, 2017

Some Watch Chickens and Others Watch Garbage Trucks

It's not just that you could robo a garbage truck; it would be relatively easy.  Software for automating heavy trucks is already almost common and the first rule of programming is don't rewrite anything.

The copilot in the truck seems to be the one calling when the truck is lined-up with a can to be collected.  S/he activates the arm and it goes through its moves.  I noticed it sometimes shook the can twice before returning it to the ground but there was no way to discern why that happened although most likely the operator controlled it.  After the process is complete, the copilot notifies the pilot, and then they go on to the next can.

That arm needs some sidewalk-facing cameras which alert the robo driver of the need to stop.  Decisions on positioning the arms are made by the minimal software needed to do it and then only servo motors are needed to make it happen.  After the process is complete, the robo collector notifies the robo driver and they continue on their robo way.

Note:  the existing trucks, in effect, have servo motors already for the arms; they're just really big ones.  To get really fancy, use stepper motors to go click, click, click to the precise position.

The robo can match the multi-shake operation seen on the meat-operated truck just now since the cans are standardized and all weigh the same.  Therefore, it only takes a weight sensor on the arm to know whether the can is actually empty.

Relative to many robo operations such as a Martian Explorer wandering around Mars which is extremely complex, picking up that can on the side of the road is just one part of the overall Martian Explorer robo toolset so it's relatively minor.  I therefore conclude, the robo garbage truck can be done and almost effortlessly using mostly existing software technology.  Most of the hardware exists already too and only needs servo motors plus some cameras to make it 21st Century.

Ed:  you will hang about watching a garbage truck just to make this point?

Yep.  It's nice out there.  Just don't step on a fire anthill or an acorn with bare feet.  That's goin' hurt.

You know, if I had the scratch for it, I believe I would be bankrolling autonomously-operated garbage trucks.  The need is clear, predictable, and unlikely to change in the foreseeable future.  I've heard of autonomous big rig trucks but no autonomous special purpose vehicles such as a garbage truck.  This won't be like buying into Apple before they announced the iPhone but it will be like buying the Utility Company in Monopoly.  It's not going to make the biggest bucks but it will make them steadily and even a small percentage in this game is a huge pile of money.

How about if you come out with the first autonomous robo garbage truck and (gasp) it actually works.  You wouldn't be able to make them fast enough when there's a whole world of demand for them and no-one else makes them.

Ed:  why shouldn't others copy you?

They will but it will take them years to catch up.  They have to tool up a production line and even with robos that's going to take time.

During the years it takes them, you become the acknowledged King of the Robo Garbage Trucks and your business is assured almost no matter what they do since there just ain't a whole lot of room for innovation beyond automating them in the first place.

One of the fundamental principles of Rockhouse job security is if you're willing to do the dirty jobs then you will never have to worry about losing one.  I was and I didn't.  This is the same thing since there's no glamour in building garbage trucks but it's still a valid service and worth big bucks.  It's not like you're going to get yer hands dirty; fookin' robos will do all the work.

No comments: