Robots doing tree maintenance

My Mom is having sleeping problems, and so I am up late.

I am going to be having Osage-orange, honey locust (probably inermis, but coppincing might bring out thorns) and black locust.  And minor things like hawthorn and fruit trees.  I think one of the big problems with Osage-orange (hedge) was maintenance.  Who wants to prune trees with such wicked thorns?  If one builds a robot, that is just a subroutine or two.  Probably more like a family of subroutines, so maybe you need Raspberry Pi instead of Arduino?

Where I think robots could work, is taking down trees.

The robot has some accurate and precise way of determining where it is, starting from when it first "mounts" the tree.  And in the general case, it is probably a family of robots, not just one.  Why?  In Canada, the coastal Douglas Fir grows to 340 feet.  In the USA, the coastal Redwood grows to 380 feet.  The Redwood has a relative; the giant Sequoia which is the largest tree on the planet.  Just a bit shorter.

I think the first robot is a "scout".  It maps the branches and climbs to the top.  Which may mean traversing bifurcations in the leader on the way up.  It''s primary objective is to determine how tall the tree is.  It will have lots of subsidiary objectives.

I think the next series of robots to go up, are cutting all the branches off.  It might be a family of robots.  The branches lower down could easily be HUGE.  The robot gets into position, and wraps a cable around itself and the tree above the branch to be cut, and tightens it to anchor itself in position.  The robot then inserts a lowering cable mount point into the branch, which has a cable going to a winch on the robot, with enough cable to lower it to the ground.

Most people think saws in cutting trees.  I am thinking chisels.  The robot drills a hole through the branch, and then positions chisels on either side (with a bolt and driving nut), and then starts to drive the nut.  And the chisels get brought in, cutting the branch off.  Once the branch is cut off, it is lowered to ground and the cable rewound.  Move up to next branch.

At some point, all the branches are gone (and this subroutine needs to deal with multiple leaders - or at least the leaders of a multi-leader tree.).  We are left with the tallest part of the tree (maybe).  These segments (ideally just 1, could be more) are all potentially sawable.  So, the robot needs a saw plan.  It goes up to the highest saw point, and slightly below that it screws itself into the tree to anchor itself.  It then drills a hole for the two chisels, and then positions the chisels.  Probably before this, it inserts an anchor for the N feet of tree above the cut point and anchor it to the winch.  The robot severs the top N foot of the tree, and lowers it to the ground.  It retracts the winch, and moves down to the next place to sever the trunk.

For lots of trees, I think pulling two opposite chisels that are scary sharp to the middle with a nut and bolt is probably the best.  But, if you get to the point where you are dealing with trees that are 10 or more feet in diameter, maybe sawing become more tractable?  I think two chisels is always going to be the least energy way of severing the trunk.

But the idea of having robots specialised in cutting branches below X height is useful, it leads to better trees to saw.  And being robots, maybe they can travel a huge area just pruning trees.  But, it is not just a single robot.  There is scout, which climbs to the top of the tree first, from which the maximum pruning height is found.

My hope, is that at some point all of this is Open Source.

To severe a trunk or branch, you are creating new surface area.  To cut (knife or chisel) only creates this minimal increase in surface area.  To saw, you are creating orders of magnitude new surface area.

Trunks are not always free of residual stress, and in cutting (saw or chisel) residual stresses can bind the cutting action.  So there can be situations where saw or saw and chisel works better than chisel alone.

They have this version of a remote control tree pruner? Is this what you are talking about?

This version is a lot faster though???

There was discussion about chisels versus saws, and in the logging industry, chisels have been used, and continued to be used quite a bit actually.

The problem with chisels, is that to cut big trees, when the chisels are forced together, it often cracks the tree. Sawmills will not accept any logs taken from a logger using these types of machines, BUT it does not mean that cracking cannot be put to good use. This machine, while a harvester, and not a pruning machine, uses 100% chisels to accomplish the task of felling trees, delimbing them, even cutting and splitting them up into firewood. I thought that was a very novel way to make firewood automated. (Skip ahead about a minute to get to the good part of the video)

Interesting robo forestry ideas Gordon. A chisel clamp does seem simpler, would probably make more cuts than a saw could before getting dull.
I've been expecting there would be all kinds of agronomic robots by now, but nothing much yet. Seems like promising products get hung up on: grant or venture capital money ran out, inventor had to get a job, garage project was difficult, life got in the way, or corporate project gets bogged down by institutional equivocation. (Like the robot not yet being able to out-compete standard equipment.)
I figure a successful project needs to get rushed into the field stage, and the builder needs to be willing to change everything 5 or 6 times to meet the actual requirements they find. None of this "oh, it worked on flat ground" or "I don't want to run it in this rainstorm" or "this arm design only works with five motors that cost $8000 each".

If I ever get a robot project going, I'd skip the computer brain - GPS - sensor integration that bogs down all the hobby projects; someone better qualified can do that part later. I'd focus on the getting the frame and kinematics perfected for the task itself.

Grady Houger wrote:Interesting robo forestry ideas Gordon. A chisel clamp does seem simpler, would probably make more cuts than a saw could before getting dull.
I've been expecting there would be all kinds of agronomic robots by now, but nothing much yet. Seems like promising products get hung up on: grant or venture capital money ran out, inventor had to get a job, garage project was difficult, life got in the way, or corporate project gets bogged down by institutional equivocation. (Like the robot not yet being able to out-compete standard equipment.)
I figure a successful project needs to get rushed into the field stage, and the builder needs to be willing to change everything 5 or 6 times to meet the actual requirements they find. None of this "oh, it worked on flat ground" or "I don't want to run it in this rainstorm" or "this arm design only works with five motors that cost $8000 each".

If I ever get a robot project going, I'd skip the computer brain - GPS - sensor integration that bogs down all the hobby projects; someone better qualified can do that part later. I'd focus on the getting the frame and kinematics perfected for the task itself.

That is what I try to do too...study the product I want to make, and then dumb it down so that it is far simpler, and can actually be made.

Like I built a feller-buncher for my logging trailer. I looked at the feller-buncher video above, and then set out making my own homemade-feller buncher. When doing a search for them, I found only one other person had made a homemade version. I think that is because people see how complex they can be and think, I could never do that.

I try to look at each machine, and break it down by steps. Broken down like that, it is not as hard to build as most people think.

Hi Gordon,

Interesting stuff.  I have been studying Arduino and Im a total novice at it still finding my bearings but I always take some time to read ahead and try to get a sense of how to use sensors to my advantage.  And looking at more advanced projects.

My goals are smallish compared to you yours, I have a rough idea for an agriculture sensor robot, testing soil moisture and temperature, air temp and weather changes, as well as being able to have eyes everywhere at the same time if need be.  I  recall a  movie when I was a kid Silent Running that had agricultural robots in it, and it always stuck in the back of my mind that such technology might someday be a reality.

I will in my spare time be on the alert for chewing and chopping and sawing robots and climbers and the sensors to make such a machine a possibility. I suppose the lazy way to do it is mount a camera and use the critter as a remote/RC controlled version with some useful behaviors at its core. But I think you are after something more.  A "smart chainsaw" that climbs like a squirrel.....

Best, Mike