Tomorrows Food Production AgRA Webinar: October 29 th 2014 Satoshi - - PowerPoint PPT Presentation

▶
tomorrow s food production
SMART_READER_LITE
LIVE PREVIEW

Tomorrows Food Production AgRA Webinar: October 29 th 2014 Satoshi - - PowerPoint PPT Presentation

Sep 22, 2022 30 likes •503 views

Automation Technology for Tomorrows Food Production AgRA Webinar: October 29 th 2014 Satoshi Yamamoto Visiting Faculty, CPAAS, WSU Senior Researcher, BRAIN, NARO 1 Motivation for the automation How to keep the current level? 140,000

slide-1
SLIDE 1

Automation Technology for Tomorrow’s Food Production

Satoshi Yamamoto

Visiting Faculty, CPAAS, WSU Senior Researcher, BRAIN, NARO

AgRA Webinar: October 29th 2014 1

slide-2
SLIDE 2

2

Motivation for the automation

  • Sustainable
  • Reliability & Safety

50,000 60,000 70,000 80,000 90,000 100,000 110,000 120,000 130,000 140,000 1920 1940 1960 1980 2000 2020 2040 2060 Population in Japan Year

*1920 – 2010: Statistics Bureau, Japan *2010 – 2060: National Institute of Population and Social Security Research, Japan

Peak: 2008

Food production

  • Efficient work
  • Information management

Automation Technology

How to keep the current level?

slide-3
SLIDE 3

TOPICS

3

  • 1. Back ground
  • 2. Components of plant factory for strawberries in

BRAIN, NARO

  • 3. 3D modeling of apple fruit in CPAAS, WSU
slide-4
SLIDE 4

National Agriculture and Food Research Organization

4

http://www.naro.affrc.go.jp/english/index.html

Researcher: 1,542 (April, 2013) The fiscal 2013budget: 529M US$ (1US$ = ¥109) Research institute under MAFF

Largest research

  • rganization addressing

“agriculture, food and rural communities”

slide-5
SLIDE 5

5

Fruit Planted Area (2012) Production Quantity (2012) Wholesale Value (2011) a) (ha) (t) (106 USD) Tomato 12,000 722,400 1,522 Strawberry 5,720 163,200 1,573 Cucumber 11,600 586,600 1,444 Egg plant 9,860 327,400 805 Sweet Peppers 3,420 145,000 602 “Unshu”, Mandarins 43,700 895,900 1,496 Apple 37,400 793,800 1,199

MAFF

a) Calculated as 1 USD = 100 JPY.

5000 10000 15000 20000 1970 1975 1980 1985 1990 1995 2000 2005 2010 Area Harvested (ha) California Japan

Outline of strawberry production in Japan

slide-6
SLIDE 6

6

Annual working hours (h/0.1ha) 2,000 Harvest season (months) 6 (December to May) Average of planted area per producer (ha) 0.3 Planting density (plants/0.1ha) 7,000 – 8,000 Production (t/0.1ha) 3 – 5

* MAFF, 2007

Seedling 10% Planting 4% Fertilization 3% Pest control 4% Cultivation management 28%

Harvesting 23% Sorting, Packing 27%

Labor management 1%

Percentage of working hours

slide-7
SLIDE 7

Plant Factory for Strawberry Production

7

  • 1. Movable bench

system

  • 2. Stationary

harvesting robot

  • 3. Sorting &

packing robot

slide-8
SLIDE 8

8

Movable bench system

 Space saving  Automated spraying  Saving energy cost  Increasing yield per area  Improvement labor condition

Movie

slide-9
SLIDE 9

9

Measurement growth information

Movable Bench System Kinect

Growth information of all plants every day

slide-10
SLIDE 10

10 Depth Image Color Image

Measurement growth information

Easy to extract leaf area using depth info

slide-11
SLIDE 11

11

4 m

  • Feb. 23
  • Mar. 29
  • Apr. 26

Color Depth Color Depth Color Depth

42 beds

Measurement growth information

slide-12
SLIDE 12

12

Lack of Iron High EC or Water stress

Health diagnosis

Measurement growth information

Basic info of plants: height & width

slide-13
SLIDE 13

13

Strawberry harvesting robot

<Development Target>

  • 1. More than 60% success rate
  • 2. 10s to pick & place a fruit
  • 3. 0.1ha / night (8-12h)
  • 4. No bruise

Prototype 1

  • Basic type (no storing function)
  • Cylindrical manipulator (3 DOF)
  • Three camera
  • Four halogen lamp
  • Finger for cutting & holding stem
  • Suction tube to cancel the depth

error

slide-14
SLIDE 14

14

Prototype 2

  • Five LED
  • Through type photo sensor
  • Tilting motion of robot hand
  • logistic function for fruit

containers Cylindrical manipulator

slide-15
SLIDE 15

15

Picking motion

Suction tube Finger Through type photo sensor a) Approach to a fruit with suction tube b) Move finger forward c) Move finger & tube backward

slide-16
SLIDE 16

16

Tiling motion before picking

a) Right direction b) Left direction Two independent air cylinders

slide-17
SLIDE 17

17

Prototype 3

  • No suction tube
  • Movable platform

Movie

slide-18
SLIDE 18

18

Prototype 4

Movie

  • One LED
  • Diffused photo sensor
  • Bending motion for placement

Shibuya Seiki Co., Ltd.

slide-19
SLIDE 19

19

Stationary type with movable bench

Movie Commercialized by Shibuya Seiki Co., Ltd.

slide-20
SLIDE 20

20

Change of robot’s faces

  • Simplicity
  • Compactness
slide-21
SLIDE 21

21

  • 1. Binarization
  • 3. Maturity assessment
  • 2. Occlusion assessment

Image processing

  • 4. Stem detection
slide-22
SLIDE 22

22

y = 0.991x - 2.7616 R² = 0.9557 20 40 60 80 100 20 40 60 80 100 Estimation (%) Human eye (%) y = 1.0633x - 1.3707 R² = 0.8207 20 40 60 80 100 20 40 60 80 100 Estimation (%) Human eye (%)

Amaotome Beni-hoppe

Difference of coloring among varieties

slide-23
SLIDE 23

23

Fruit condition

10 20 30 40 50 60 70 80 90 100 Aisle (Feb.) Bed (Feb.) Fruit condition (%) ‘Beni hoppe’ cultivar Aisle (May) Bed (May)

A B C D E Bed Side Aisle Side

A B C D E

slide-24
SLIDE 24

24

Harvesting from bed side

Prototype 2 Prototype 3 Mayekawa mfg. Co., Ltd. Waseda University

slide-25
SLIDE 25

25

Harvesting from bed side

Hand-eye-camera for stem detection Stereo vision for position detection Movie

slide-26
SLIDE 26

26

Reduction of influence of fruit condition

Movie

Separate from adjoining fruits Approach Pick Place

slide-27
SLIDE 27

27

Reduction of influence of fruit condition

End-effector a) Vacuum b) Grip Mobile bench Unit 7 DOF Manipulator Coloration Measurement Unit Position Detection Unit

Movie

slide-28
SLIDE 28

28

Reduction of influence of fruit condition

Movie 3DOF Manipulator Mobile Bench Unit 7DOF Manipulator Vacuum Hand Picking Hand Coloration Measurement Unit Position Detection Unit

slide-29
SLIDE 29

29

Mini-summary for harvesting robot

2003 2013 Stationary harvesting robot Harvesting success rate: 40 – 70 % 2010 2006

  • Cylindrical manipulator (3 DOF)
  • Three camera, Four halogen lamp
  • Finger for cutting & holding stem
  • Suction tube
  • Machine vision & software: Maturity, Occlusion…
  • Finger shape
  • Tilt function of robot hand
  • Diffused photo sensor

Don’t move expensive robot!

slide-30
SLIDE 30

30

Strawberry sorting & packing robot

From harvesting box to shipping tray

slide-31
SLIDE 31

31 Single layer 1 Double layer Small pack Single layer2 Single layer 3 Hart shape

Shipping types

slide-32
SLIDE 32

32

Strawberry sorting & packing robot

Supply unit Sorting & Packing unit

Movie Single-layer tray Returnable tray

slide-33
SLIDE 33

33

Supplying unit

Camera Manipulator (3 DOF) Suction hand Harvesting container

slide-34
SLIDE 34

34

Sorting & packing unit

Collision Safe

Camera Manipulator (4 DOF) Suction hand

Single-layer tray Returnable tray

slide-35
SLIDE 35

35

Strawberry sorting & packing robot (2)

Machine vision: Kinect Conveyer for harvesting containers Fruit conveyer Conveyers for shipping trays Machine vision: Color camera End-effector Manipulator

slide-36
SLIDE 36

36

Start Supply fruits and shipping tray Detect the suction point of target fruit in harvesting container Pick up fruit, move to digital camera Weight and orientation of the held fruit Place on shipping tray Stop Continue?

Kinect Digital camera

slide-37
SLIDE 37

37

Segmentation of fruits in harvesting container

Segmentation of fruits using color & depth info

slide-38
SLIDE 38

38

Fruit orientation

Size & Orientation V of HSV R – G image Movie Maximum error: 25.1˚ MEAN : 0.3˚ SD: 5.1˚

slide-39
SLIDE 39

39

Strawberry packing robot in grading line

Movie

Packing Robot

IR sensor Weight scale Yanmar Green System Co., Ltd. Color camera

slide-40
SLIDE 40

40

Strawberry packing robot (Basic)

slide-41
SLIDE 41

41

Mini-summary for sorting & packing robot

2013 2011 2007 Robot hand Supplying unit Packing robot (Basic) Sorting & Packing robot using Kinect Packing robot in grading line 7 s / fruit 4 s / fruit 1.5 s / fruit < 1 s / fruit

More than human ability!

slide-42
SLIDE 42

42

3D modeling of apple fruit in CPAAS, WSU

  • Density: important factor for evaluation of a fruit inner

quality.

  • Volume: not a common technique in a fruit sorting system.
  • Appearance: check using surface color information.

3D reconstruction for fruit sorting system using Kinect

slide-43
SLIDE 43

43

Measurement setting

Kinect Apple LED

slide-44
SLIDE 44

44

3D models using Kinect

How should we use them?

CAD data can be download from website of GrabCAD.

Automated grading system

  • Inner quality: density
  • Appearance assessment
slide-45
SLIDE 45

45

Summery

Packing robot in grading line Grading based

  • n 3D model

Stationary harvesting robot Growth measurement Movable bench system

slide-46
SLIDE 46

46

For the tomorrow’s food production…

  • Is it time to get out of the plant factory?
  • Construction the preferred environment for automation will be important.
  • Consumer 3D sensor has changed the accessibility to 3D info..

Stem detection will be a key for a robotic harvester… Over the Row Sensor Platform (left), Detection of apple fruits (right) CPAAS, WSU (Prof. Karkee) Simple hardware & smart software

slide-47
SLIDE 47

47

Thank you for your attention!