Hacker News Reader

Comment by Loughla | Hacker News Reader

I can't get your site to load, so I checked out the hackaday page.

Trying not to be overly cynical here, but how is this worth the cost? It appears that it simply plants, waters, and detects/removes weeds. In a 1,250sq.ft. garden, we invest less than 2 hours per week on these tasks.

How would this be scalable? How do you spell scalable?

How would this justify its cost?

How does it withstand being outside all year, year round?

How does it not just destroy your crops when they grow tall?

How could this possibly improve on current farming methods (outside of removing chemical weed-killers maybe)?

I understand it's in its infancy, but I'm genuinely having a hard time with this.

View on HN

> How would this be scalable?

Extend the x and y axis?

> How would this justify its cost?

Presumably by saving costs elsewhere? Labor, chemical inputs. Not saying it's there yet though, but why should it be?

> How does it withstand being outside all year, year round?

A cursory glance at the specs indicates that they've considered waterproofing at least. I imagine "the elements" will prove even more challenging than expected, but it's not like this wasn't considered.

> How does it not just destroy your crops when they grow tall?

The y axis is up pretty tall, and the manipulator is pretty narrow to get down in between plants. Clearly it is not advisable to grow crops that are taller than will fit in the machine.

> How could this possibly improve on current farming methods

Automation? We don't expect Burger King robots to improve on culinary techniques. I feel this is mostly interesting for repeatable research uses, but as basic manual labor savings it's interesting as well.

(I am in no way affiliated with this project... just seems easy to see the potential, and odd to treat a prototype as if it needs to make a business case right out of the gate)

I'm genuinely not trying to treat it as if it's a business right away. It's just, I don't know, maybe I'm too close to farming to get my head out of it.

I'm a 4th generation farmer in my spare time. The variables required in agriculture are just too many for something like this, is my opinion. I get it, farming isn't immune from technology, but this just seems like over-engineering a problem.

> The variables required in agriculture are just too many for something like this, is my opinion.

I'm certain there is a lot of complex knowledge locked up in farmers' brains that most folks wouldn't even think of. The same is true for doctors though, and it doesn't seem foolish that folks are creating expert systems to help with complicated diagnostics and such. Surely the wisdom of farmers can be codified and emulated reasonably well by software.

Further, maybe this is only limited to a subset of agriculture. It doesn't need to apply to every crop type to be useful. It's probably dumb to grow 100 acres of wheat this way. Perhaps it's useful for highly-controlled small beds of high-value high-labor-input vegetables or such?

I would love to see a shared open database of environment and care parameters. Imagine all of the ounces of water per week per watt of sunlight type optimizations. Identifying okra that is ready to pick would be nice -- the maximum size for tenderness and plant health. A shared open db of farming parameters would be awesome.

extending the x and y axis is not a solution. Complexity and costs will go up non-linearly with such a solution. A better solution would probably involve wheels... Then they could call it a "tractor." (but seriously, that'd be a better solution than a 100 meter (or even 3 meter) x/y axis.

I think better idea would be four tall legs (like a table legs) so it can move around the field. It doesn't need to move fast. You also need something to detect its position (laser beacons? radio beacons?) and some place to hide from rain. And a hand with a camera for harvesting crops into a basket.

I completely agree with you in its current incarnation. There is no way this is even within an order of magnitude of the price/performance ratio of a migrant farm worker (who are career farmers and much faster than you or I).

However, I suspect a device like this will scale well cost wise with longer rails, thus larger versions working in a elongated greenhouse would probably significant efficiency. Additionally, you will need a device like this if you want to do vertical hydroponic farming, as the efficiency of people drops precipitously as soon as they can no longer easily reach the plants they're working with.

Having worked in a greenhouse, much of the planting is automated in large-scale shops already through vacuum planters. Watering is similarly automated through irrigation systems.

Now, vertical hydroponics, that's one I hadn't considered. That may be a good point. I lack any experience with those systems, though, so I hesitate to comment.

Might be cool if you can stick it on the back of a tractor and have it do these things while in motion.

Even with auto-steer technology and GPS, I can't see how it would be accurate enough to be useful over any size field.

You would spend more machine/human time scanning and driving, more money in fuel, and more energy in general than if you were to just plant, spray, and irrigate.

Reply on news.ycombinator.com