Agriculture robots - what doesn’t exist but should?

[Thatmuppetoverthere]

Hi all,

This is probably going to be one of those weird threads.

So first, a bit of context. I’m an electrical engineering/robotics student here in the UK. As a part of my studies I have to design and ultimately build a robot that has real world applications. Most of my classmates are building somewhat silly robots or industrial-style robots that’d cost a fortune.

Personally, I’m a little more practical and (perhaps somewhat naively) want to make the world a better place through the technology I will build. And I’ve chosen farming for that, as without farmers who make the food that feeds me, my degree is worth less than nothing.

I’ve taken a look at the agricultural robots currently on the market, and those that do exist are incredibly good… but astronomically expensive. Far beyond what most non-corporate farmers can ever afford (especially in light of the recent protests which are partly what set me on this course), giving those mega farms an even greater advantage over normal hardworking farmers.

I recently came across the grain weevil, which uses a robot in place of humans in grain bins/silos. I had initially set upon improving that design, however now understand that grain bins are very rare here in the UK and thus, my efforts would be mostly wasted.

My question to this forum is thus: What part of farming is most dangerous here in the UK and would best be replaced by a robot? And what would be a price for such a robot that would be economical for most farmers?

I look forward to your guys’ help and ultimately putting that robot into action!

- Alex

PS: I’m not all that fussed about what area of farming it is; I live in Lincolnshire so there’s plenty of farms I can use for reference/study and have most of the summer to come up with the design.

 

[sheep boy 3000]

Drones.
Drones for spraying bracken which do exist but the HSE would rather focus on theoretical risks rather than the real risks on quad bike users not being able to see where they drive or the massive increase in ticks. I did lambs today that had 20-30 on their faces and I also took one off my dog. Lymes disease is the silent epidemic.
Robots to shear sheep would be good but understandabley difficult.
Robots for fencing. Positioning stobs accurately and laying out materials as they go.
There are many opportunities.
Something as simple as my robotic vacuum cleaner has already been a massive boost personally. Time is tight and easier days dont need taken up with an afternoon with Henry.
Good luck

 

[PSQ]

We have had some amazing technology hit the mainstream in farming over the last 20 years. Automation, self steering high precision GPS, variable rate application of seed and fertilisers, scanning and mapping of soils (Hutchison's Omnia), remote sensing by drone or satellite of many parameters including NDVI, yield mapping and now protein mapping using Xray florescence mounted on the combine (JD's Harvest Lab 3000), etc.

Even with all of the new tech I live for the days when I'm sitting on a machine optimising whatever process I'm controlling, and knowing that it's running 'just so'. The problem is we can all see what's coming down the road, as the next area of improvement is going to be to replace the soft squishy bit in the field with an algorithm that can asses all variables more efficiently and optimise all parameters 100 times per second, that can work 24/7 and doesn't need to take days off or adhere to ever more stringent working time directives. We'll keep calling it progress, right up to the point that we read the service agreement with CNH or JD or whoever, and then realise that we're no longer allowed to be within a kilometre of the farm and 'our machines' due to 'health and safety concerns'. But when we are no longer needed, then we will no longer be needed, and land ownership will transfer to the corporations and the venture capitalists who pull their strings.

I thought the dystopian nightmare was a few decades away as all of the attempts to automate field operations look like a bodge job on a traditional tractor or a obvious false start from the likes of Nexat, or so niche like little solar powered robot weeders. Then amongst the usual tractors covered with sensors from Claas, JD, CNH etc at Agritechnica I spotted something that looks like a realistic workable prototype, and realised that the future has arrived a bit quicker than I was hoping:

So OP, whatever you design please be kind to the tractor drivers and the operators, as we are led out to pasture just like the cart horses that slowly faded from the landscape at the start of this adventure 100 years ago.

 

[BrianV]

With the shortage of pickers for vegetables & flowers there would seem to be a market there

 

[Chae1]

Something that identifies the Dock weed in grassland and sprays them.

 

[white_stuffed]

Pressure washing robot.

 

[Muddyroads]

For me, it would be a weeding robot for vegetable production, particularly if it could be priced at a level that would be justifiable for smaller growers before we all go extinct. I have in mind something like the robot mowers, but probably using an electric current to kill the weeds. Ideally it could “learn” what each crop is in order to be able to weed between plants, not just between rows.

 

[DrWazzock]

A roguing robot to pull out and cart off blackgrass plants.
In theory it’s easy. Locate centre of plant, lower a “hand” down which rotates and tightens round it, pulls it out and dumps it in a hopper. Run on 4 bike wheels that go between the cereal rows. At lot of emphasis would be on the software that recognises and locks on to the plant. The “hand” would need a bit of mechanical design. Maybe pneumatic hydraulic. Might already exist as modules that need integrating.
Cost? Don’t know. Glad you are working on stuff like this. Keep at it.

 

[TelesnaAg]

For me it would be a sex doll real life cant tell the difference robot chick,a 11 out a 10....who i can turn off when things are busy.

Perfect.

Ant...

 

[melted welly]

Address the labour shortage and its resultant cost to producers, so veg pickers, graders, weeders, sprayers etc.
Single bed/row machines that will work 24/7 and affordable to smaller producers. With agency labour at >£14/hr and >£20/hr overtime, payback could be quite quick if successful.

 

[AndrewM]

dont bother with drone spraying, there arent any pesticide products licensed for application

as student project you cant get to ambitious if you have to build it.

that grain weevil robot, could you make one that works like a mole or worm, that can go through the grain pile instead of ontop of it? could be useful in grain sheds to hunt out hot spots that need a fan placing in them.

are there any good robots for cleaning out industrial roof gutters, or even better roof valleys?

anygood at ai programming? can you modify a robot lawn mower with a camera to just go mow the thistles/docks in grass field but leave the grass ? maybe add a spot sprayer?

 

[Paddington]

Automated thistle puller which can trundle around the fields all day without supervision. We have one already, a version 1.1 but its fingers are getting stiff and its wellies leak.

 

[Clive]

I want a tiny robot(s) that can mulch weeds, companions and N fixing plants between 25cm wheat rows ...... I want them to run and recharge autonomously from easy moved field based solar powered battery pack charge stations 24/7 and I want them to cost no more than a DJI drone so I can buy dozens of them

this could replace need for (or at least reduce significantly) herbicides and synthetic N P K in combinable crops - that makes the world (climate and biodiversity) a much better place and improves peoples nutrition whilst helping farmers become less dependant upon input suppliers


Happy to work with anyone that wants to make this realty .......... how hard can it be ?

 

[puppet]

For me, it would be a weeding robot for vegetable production, particularly if it could be priced at a level that would be justifiable for smaller growers before we all go extinct. I have in mind something like the robot mowers, but probably using an electric current to kill the weeds. Ideally it could “learn” what each crop is in order to be able to weed between plants, not just between rows.

An electric weeder is already in development. I went to a presentation on it.

 

[sheep boy 3000]

Something I would buy that an engineering student could probably do on their lunch break is a dosing gun for livestock that can administer 2-3 different drench's in one go.
The different dose rates would be better measured out further back at the container as well as the power source, this would leave the dosing gun more compact than the Te-Pari.
The Te-Pari is very bulky and the nozzle isnt where it should be in relation to your hand and it only administers 1 drench but through a powered gun. Normally it administers 1 drench before you go back to a manual gun because it us unreliable.
It might not be robotics or glamorous but I guarantee if you get it right it would sell around the world. Put a bit of effort into the design of the container(s) so it is comfortable and easy to put on and take off rather than the historically poor design of 2 or 1 strap that never sits right.
One other thing if you can build it is dont give it to a pal to test but trial it on larger units. Shearing contractors or livestock markets will be able to point you in the right direction

 

[exmoor dave]

Something I would buy that an engineering student could probably do on their lunch break is a dosing gun for livestock that can administer 2-3 different drench's in one go.
The different dose rates would be better measured out further back at the container as well as the power source, this would leave the dosing gun more compact than the Te-Pari.
The Te-Pari is very bulky and the nozzle isnt where it should be in relation to your hand and it only administers 1 drench but through a powered gun. Normally it administers 1 drench before you go back to a manual gun because it us unreliable.
It might not be robotics or glamorous but I guarantee if you get it right it would sell around the world. Put a bit of effort into the design of the container(s) so it is comfortable and easy to put on and take off rather than the historically poor design of 2 or 1 strap that never sits right.
One other thing if you can build it is dont give it to a pal to test but trial it on larger units. Shearing contractors or livestock markets will be able to point you in the right direction

Great Idea,

Even if it were just 2 separate doses so you could do a wormer and a mineral drench in one pass

 

[Welderloon]

For me it would be a sex doll real life cant tell the difference robot chick,a 11 out a 10....who i can turn off when things are busy.

Perfect.

Ant...

Already done, Google - Erica, Japanese AI sex Robot hopefully version 36 will be able to drive tractors, chase cows & pass her welding codes

 

[TelesnaAg]

Already done, Google - Erica, Japanese AI sex Robot hopefully version 36 will be able to drive tractors, chase cows & pass her welding codes

I knew this AI shite would come in handy oneday!

Ant...

 

[Bury the Trash]

Something I would buy that an engineering student could probably do on their lunch break is a dosing gun for livestock that can administer 2-3 different drench's in one go.
The different dose rates would be better measured out further back at the container as well as the power source, this would leave the dosing gun more compact than the Te-Pari.
The Te-Pari is very bulky and the nozzle isnt where it should be in relation to your hand and it only administers 1 drench but through a powered gun. Normally it administers 1 drench before you go back to a manual gun because it us unreliable.
It might not be robotics or glamorous but I guarantee if you get it right it would sell around the world. Put a bit of effort into the design of the container(s) so it is comfortable and easy to put on and take off rather than the historically poor design of 2 or 1 strap that never sits right.
One other thing if you can build it is dont give it to a pal to test but trial it on larger units. Shearing contractors or livestock markets will be able to point you in the right direction

tepari gun is good for pouron. and large amounts generally but yes all guns need a redesign and shoukld be aloing the lines of the zolvix gun which is the fastest most natural to use one ever made , but with a bigger capcitiy,
the patent for it needs releasing

the old vetrazin back pack sling was /is a good bottle sling if theres any still around. and not now in shreds...

 

[deleted user 837354]

Who in their right minds would choose agriculture as an intended marketplace for a high tech robotic product? It's so low value compared to other industries. Agriculture would be the very last place I would try to sell into if it was me but I'm not the robot engineer.

My first starting point would be a robot than can seek out ground-scattered mines. Defence contracts are where the real cheese is to be had.