Author Topic: Robotic Apple Picker  (Read 3090 times)

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Oscar

zands

  • mango_zango
  • Hero Member
  • *****
  • Posts: 4455
    • Zone 10b, Florida, USA, 33321
    • View Profile
Re: Robotic Apple Picker
« Reply #1 on: May 16, 2017, 07:17:12 PM »
http://futurism.com/the-next-step-in-automation-a-robot-that-picks-apples/
The growers and the agricultural universities are always working on these robots and machines. Apples are shipped to market under ripe as it is. With robot apple pickers the under ripe situation will get worse because they will try to pick all (or 90%) apples off the tree  in one pass of the machine.

Of course not all apples ripen on a tree at the same time so the robot/machine picker will lead to lower quality apples in the stores. On the positive side I like the new common variety apples we see these days like Gala, Fuji, Pink Lady and more.



Pink Lady tree
« Last Edit: May 16, 2017, 07:21:40 PM by zands »

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Re: Robotic Apple Picker
« Reply #2 on: May 16, 2017, 10:07:36 PM »
You're probably right about quality going down. But this type of robot is a lot more sophisticated than say the robots that picked square tomatoes 35 years ago. Those would just make one pass, pull out the whole tomato plant, and shake vigorously. The tomatoes had to be bred to be super hard to withstand that treatment. Those tomatoes were so tasteless that they were discontiued. These are regular apples. But the apples shown were planted on trellis. I think most apple orchards don't have apples planted on trellis. I wonder if this robot could work on regular apple trees?
Oscar

fyliu

  • Hero Member
  • *****
  • Posts: 3216
    • Burbank/Covina, CA 10a
    • View Profile
Re: Robotic Apple Picker
« Reply #3 on: May 17, 2017, 02:03:52 PM »
They're using computer vision so they can at least see the red apples are ripe. Some people were trying to develop a system for self checkout lines at supermarkets to recognize the types of fruits a few years ago but we haven't heard much about it recently. There's also some reflectivity index that the robots could use to determine the ripeness of green fruits I think. Maybe they can even take brix readings of fruits like how the Japanese machine sorts mango by their brix and size without damaging the fruits. They should also be able to determine the hardness of the fruits by doing something like the air blow test they do at optometrist offices. That way they pick fruits that are still shippable but not too unripe.

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Re: Robotic Apple Picker
« Reply #4 on: May 17, 2017, 05:26:43 PM »
They're using computer vision so they can at least see the red apples are ripe. Some people were trying to develop a system for self checkout lines at supermarkets to recognize the types of fruits a few years ago but we haven't heard much about it recently. There's also some reflectivity index that the robots could use to determine the ripeness of green fruits I think. Maybe they can even take brix readings of fruits like how the Japanese machine sorts mango by their brix and size without damaging the fruits. They should also be able to determine the hardness of the fruits by doing something like the air blow test they do at optometrist offices. That way they pick fruits that are still shippable but not too unripe.
How can you determine brix without cutting into the fruit?
Oscar

Tropicdude

  • Hero Member
  • *****
  • Posts: 2117
    • Broward County, Florida, USA
    • View Profile
Re: Robotic Apple Picker
« Reply #5 on: May 17, 2017, 06:24:24 PM »
I guess someday we will have machines that have sensors such as spectrometers, and molecular sensors, that basically sniff, and will "know" when the fruit is at its best for picking.

This system doing that simple up and down scan seems slow and inefficient.   think of a machine that in one image wide shot, pre-maps out  the fruit coordinates for the robot sniffers to directly to them, and see if they are ready,  if not, computers estimate how much time is needed for that particular fruit. and will return in next time.

Sensors,  Fuzzy logic ( AI )  will make these things practical.  and I do not believe this is far off.  actually I think we have the tech now.  just needs to be applied. and made practical.

Sniffer tech for mango
https://www.sciencedaily.com/releases/2016/05/160510093910.htm

Spectral Sensor:
Fruit Quality Evaluation Using Spectroscopy Technology: A Review
www.mdpi.com/1424-8220/15/5/11889/pdf

There are  a few companies already developing "smart" robots for crops.  pieces are falling into place,     heck who knows we might have smart drone pickers, that go out, pick fruit, deliver it to storage, and fly to their own charging docking station when they need to, and then go back to work.   one computer could control hundreds of these.   maybe even have small zappers for bugs, keep a eye on disease, even minor pruning. flower pollination, etc.
« Last Edit: May 17, 2017, 06:36:53 PM by Tropicdude »
William
" The best time to plant a tree is 20 years ago.....The second best time, is now ! "

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Re: Robotic Apple Picker
« Reply #6 on: May 17, 2017, 06:55:01 PM »
It's all possible right now. The only question is the cost? If a really sophisticated robots costs $5 million, then how many apples would it have to pick to be cost efficient? I think in time the costs of the robots will go down and it will happen. Was all predicted in early sci-fi movies, like Silent Running.
Oscar

fyliu

  • Hero Member
  • *****
  • Posts: 3216
    • Burbank/Covina, CA 10a
    • View Profile
Re: Robotic Apple Picker
« Reply #7 on: May 17, 2017, 10:29:13 PM »
I have no idea how the brix machine works. I saw it in a youtube video where the older shirtless Japanese grower, his apprentice and the guest picked some mangos from the greenhouse and then brought them in to go through the machine that the sorted them automatically. One of those egg of the sun videos. https://www.youtube.com/watch?v=KVr5zQc0F8U

Yes, cost and scale. It takes time to miniaturize all the sensors. Once they develop the full scale ones, I imagine they'll have to sell enough to recover their money before doing the next steps.

echinopora

  • Member
  • ***
  • Posts: 231
    • View Profile
Re: Robotic Apple Picker
« Reply #8 on: May 18, 2017, 05:04:08 AM »
http://www.abc.net.au/news/rural/2015-11-02/mango-quality-in-firing-line-with-near-infrared-scanning-gun/6903024
The thing about a robot is there is a large capital investment,  but low operating costs. Multiple picks wouldn't add much to cost like it would with human labor.

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Re: Robotic Apple Picker
« Reply #9 on: May 18, 2017, 06:43:32 AM »
http://www.abc.net.au/news/rural/2015-11-02/mango-quality-in-firing-line-with-near-infrared-scanning-gun/6903024
The thing about a robot is there is a large capital investment,  but low operating costs. Multiple picks wouldn't add much to cost like it would with human labor.

Yes, so it's only good for extremely large operations. Either that or as a rent out only unit during harvest season.
If a robot costs only $1 million, and labor cost is a high $20/hr.  you still would be able to hire workers for 50,000 hours. That's a lot of repeat picking.
Oscar

echinopora

  • Member
  • ***
  • Posts: 231
    • View Profile
Re: Robotic Apple Picker
« Reply #10 on: May 18, 2017, 06:08:09 PM »
Long post but once I started looking into it I found it fairly interesting. Apples are bigger business than I thought.

It would definitely be suited to a particular type of operation. If implemented it could do to the small orchard what modern thresher/combine units did to small grain farmers. If the same machine could be tooled to harvest several fruit crops (say mango, citrus, apples, avocado) then 1 mil would be cheap. Think of australia. Broad acre fruit farming is common here, generally on fairly flat terrain. Trellised apple orchards are more the rule than the exception. Our labor costs are high. 40%-60% of the crop is wastage because the local market can only handle so much b grade fruit, and the price for low grade fruit is too low to consider picking and shipping for processing. Last time I was in western Australia for the apple harvest there were piles of low grade fruit double overhead the size of tennis courts everywhere. If the automated picker could decrease fruit damage by getting them sorted and on trays without risking crush damage in bins, that would increase export quality volume right away. This is where the industry wants to go because the local market is saturated. I never realized how many apples the world grows but have a look at this.  https://en.m.wikipedia.org/wiki/List_of_countries_by_apple_production . I think there is volume for mass production of robots. Can that china figure be right?

Between mango, apple and citrus that's year round picking. If you run these units like the automated harvesting companies do on broad acre soy/rice/corn/grain.
1mil @ 6.5% over 10 years about 140k a year plus 5% equipment value for parts and repair and 5k insurance = 195k a year
Pick 250 days, 20 hrs a day. 105 days travel and maintenance = 50000 man hour equivalents.

That prices a machine hour at about 4 dollars. Write off interest costs in the business,  plus write offs on the depreciation schedule for plant and equipment. If you increase export quality fruit volume and cut wastage, crop volume and grade will increase. You could probably make your margin on the decreased wastage and increased export volume alone. That still leaves an 11 dollar gap assuming $15/hr for migrant labour. Early adopters would make some cash, then I imagine apple prices would crater leaving only the automated producers at a new lower floor price and normal business margins.

Long post but the ROI on equipment in my business is a fraction of that, but it still pays the bills. Automation is a threat in my line of work, so usually pretty interested in where it's at. Question is will it really be an improvement when all the unskilled jobs go overseas or to robots and the locals try the fill the gap with consumer credit and mortgage debt.


« Last Edit: June 15, 2017, 08:06:27 PM by echinopora »

fruitlovers

  • Hero Member
  • *****
  • Posts: 15883
  • www.fruitlovers.com
    • USA, Big Island, East Hawaii, Zone 13a
    • View Profile
    • Fruit Lover's Nursery
Re: Robotic Apple Picker
« Reply #11 on: May 18, 2017, 10:58:37 PM »
There is nothing in this report about making the robotic picker be able to work on different fruits. Ofcourse that would make it a lot more useful, and also a whole lot more compicated, and so the cost would skyrocket.
I don't think this kind of picker would be suitable for small orchards, only mega plantings.
Yes apples are a large crop, one of the most cultivated fruits.
Oscar

echinopora

  • Member
  • ***
  • Posts: 231
    • View Profile
Re: Robotic Apple Picker
« Reply #12 on: May 19, 2017, 05:32:32 AM »
Sorry to offend Oscar, just got excited over the possible implications. I grew up in an agricultural area that does mostly canola and alfalfa. Its all automated now and there is no such thing as a small farm anymore. If this technology becomes viable and only suits mega farms, I imagine mega farms will be all that remains. If that pushes apple prices down, they'll just find more things to put apples into like they did with corn and canola.