Post on 22-Mar-2018
2/9/2015
1
Is Robotic Milking Right for
You?Jim Salfer1 and Marcia Endres2
1U of MN Extension, St. Cloud2Department of Animal Science, St. Paul
University of Minnesota
salfe001@umn.edu
© 2014 Regents of the University of Minnesota. All rights reserved.
New Seat Belt law
This becomes
effective Sept 1, 2015The National Highway Safety Council has done
extensive testing on a newly designed seat belt.
Results show that accidents can be
reduced by as much as 95%
when the belt is properly installed.
2/9/2015
2
© 2014 Regents of the University of Minnesota. All rights reserved.
First Robot Milker (1981)
Courtesy of DeLaval
© 2014 Regents of the University of Minnesota. All rights reserved.4
Milking robots are here to stay!
– North American Data
• >2500 AMS units
• >1000 farms
• >140,000 cows
• >381,000 milkings/d
• Avg 2.5 AMS units/farmRodriguez, 2014
2/9/2015
3
© 2014 Regents of the University of Minnesota. All rights reserved.
Why robotic/automated milking?
• Improve lifestyle
• Labor management
• Human health
• Latest technology
© 2014 Regents of the University of Minnesota. All rights reserved.
• Improved product quality and consistency
• Reduce direct and overhead cost
• Improved accuracy and repeatability (cows like consistency)
• Improved quality of work for employees (higher skilled workers)
• Improved workplace health and safety (take over unpleasant, arduous or health threatening tasks)
• Reduced waste and increased yield
• Reduced turnover and recruitment difficulty
Why Industries Invest in Robots
ABB – ten reasons to invest in robots
2/9/2015
4
© 2014 Regents of the University of Minnesota. All rights reserved.
Labor on a Dairy Farm
Business management
50-60%
Feeding 10-20%
Reproduction & Health 5-15%
Milk harvesting 5-10%
Milk harvesting 50-60%
or managing labor
Feeding 10-20%
Reproduction & Health 5-15%
Business management 5-10%
Conventional milking Automatic milking
© 2014 Regents of the University of Minnesota. All rights reserved.
Potential AMS Advantages
Provides Data• Milk production etc
• Over 100 measurements at every milking
• Timely decision making
Other benefits:• Consistent milking routine
• Higher skilled labor
• Never late for work
• Never needs training
• Doesn’t need scheduling or holidays off
2/9/2015
5
© 2014 Regents of the University of Minnesota. All rights reserved.
• “Plug and play,” “Plug and pray,” or
“Plug and pay”
• Low Return on Investment? (compared to what?)
• Obsolescence
• Repair costs
Potential challenges
Adapted from Bewley, 2013
© 2014 Regents of the University of Minnesota. All rights reserved.
Automated/Robotic Milking Systems
Box systems• Lely
• DeLaval
• GEA Farm
Technologies
• AMS-Galaxy
• BouMatic Robotics
Parlor systems• GEA Farm Technologies
• DeLaval
• MiRobot
• BouMatic Robotics
What is the right fit for me?
2/9/2015
6
© 2014 Regents of the University of Minnesota. All rights reserved.
Single Box
Systems
© 2014 Regents of the University of Minnesota. All rights reserved.
Multiple-Box Systems
2/9/2015
7
© 2014 Regents of the University of Minnesota. All rights reserved.
GEA Robot for Rotary Parlors
DairyProQFully Automated
Industrial Milking
Slide compliments Greg Larson, GEA
GEA Apollo – semi automated milking
© 2014 Regents of the University of Minnesota. All rights reserved.
DeLaval Rotary Parlor
Slide compliments Mark Futcher
2/9/2015
8
© 2014 Regents of the University of Minnesota. All rights reserved.
MiRobot version 1.0 under cow model Slide compliments David Rubin, Microbot
MiRobot
© 2014 Regents of the University of Minnesota. All rights reserved.
You must like working with cows
• “Management makes milk – Robots
only harvest it” Doug Kastenschmidt – Ripon WI
• Cow management must still come
first
Keys to success with robots
2/9/2015
9
© 2014 Regents of the University of Minnesota. All rights reserved.
Guidelines for Efficiency
Box robots
140-190 attaches/24 hrs
2.4-3.0 milkings/cow/day
Goals for milk per robot
• 4000-4500 lbs – OK
• 4500-5000 – Good
• >5000 - Excellent
© 2014 Regents of the University of Minnesota. All rights reserved.
Pounds of milk per robot
0
1000
2000
3000
4000
5000
6000
Average = 4,325
2/9/2015
10
© 2014 Regents of the University of Minnesota. All rights reserved.
Number of milkings per day
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Average = 2.6
© 2014 Regents of the University of Minnesota. All rights reserved.
Excellent feed management
Survey results
– Feed management ranked 1st
– Pellet palatability and quality ranked 2nd
Nutritionist that like the challenge
of robots
Keys to success with robots
2/9/2015
11
© 2014 Regents of the University of Minnesota. All rights reserved.
Goal of every feeding program
1. Meet nutritional needs of cows while maintaining
cow health
2. Optimizing milk and components
3. Economical
4. Labor efficient and cost effective feed delivery
system
© 2014 Regents of the University of Minnesota. All rights reserved.
Automatic milking system feeding
Additional AMS goal:
• Entice cows to visit the milking station regularly and
frequently
Accomplished through:
• Partial mixed ration (PMR) at bunk
• Additional concentrate at the milking station
• Series of selection gates in a guided flow system
2/9/2015
12
© 2014 Regents of the University of Minnesota. All rights reserved.
Feeding to meet the farmers
goalsBalancing energy in AMS vs. bunk
• high forage/low energy TMR drives cows to robot – may limit milk production
• high energy TMR – increase late lactation “lazy” cows
How much concentrate to feed through AMS
• increase risk of acidosis and lameness
• off feed problems
0
5
10
15
20
25
0.7
0.7
25
0.7
33
0.7
41
0.7
48
0.7
53
0.7
57
0.7
62
Perc
ent
laz
y co
ws
Bunk ration energy, Mcal Nel/lb
Lazy…
Rodenburg, 2008
© 2014 Regents of the University of Minnesota. All rights reserved.
Daily robot concentrate
Free flow Guided flow
Average, lbs 11.2 7.9
Minimum, lbs 2.0 2.0
Maximum, lbs 25.0 18.0
2/9/2015
13
© 2014 Regents of the University of Minnesota. All rights reserved.
PMR balance vs milk tank average1
Free flow Guided flow
Average 20.0 12.3
Minimum 10.0 9.0
Maximum 30.0 20.0
1Bulk tank average milk – PMR balanced milk
© 2014 Regents of the University of Minnesota. All rights reserved.
Voluntary and Involuntary milkings of
cows fed at two concentrate levels
6.6 lbs
AMS pellet
17.6 lbs
AMS pelletP-value
Total milking/d 2.6 2.8 .13
Not fetch cows 2.4 2.7 <.05
Bach, JDS 2007
2/9/2015
14
© 2014 Regents of the University of Minnesota. All rights reserved.
Milking frequency of high starch
vs. low starch pellets1
23%
46%
27%
4%
High Starch Pellet2
2 X
3 X
4 X
5 X
35%
38%
15%
12%
Low Starch Pellet3
2 X
3 X
4 X
5 X
1 Avg consumption – 11.9 lbs of pellets. 249% starchy grain, 325% starchy grain. Halachmi, et al, JDS 2006
Average:
3.39 visits
78 lbs
Average:
3.31 visits
79 lbs
© 2014 Regents of the University of Minnesota. All rights reserved.
Pellets vs Meal
Pellet vs meal effect on visits
Item Meal1 Quality pellet
Visits/cow/day 3.57 3.93
Milkings/cow/day 2.35 2.50
Milk/cow/day, lbs 53.6 57.2
Rodenberg, 20041 49% distillers grains, 49% cracked
corn, 2% molasses, 0.1% flavoring
2/9/2015
15
© 2014 Regents of the University of Minnesota. All rights reserved.
High vs low quality pellet
Pellet quality effect on visits
Item Low Quality High Quality
Visits/cow/day 3.40 4.04
Milkings/cow/day 1.72 2.06
Fetched cows, % 16.0 7.1
Milk/cow/day, lbs 54.5 55.6
Rodenberg, 2002
© 2014 Regents of the University of Minnesota. All rights reserved.
Feed Costs of Robot vs. Conventional
Total Feed Cost
Production Parlor Robot
120 lbs $8.87 $9.40
90 lbs $7.53 $7.85
70 lbs $6.80 $6.50
50 lbs $5.63 $5.18
AVG $7.21 $7.23
**Ration data from 5/1/12
(CS-$50/ton, Hlg-$60/ton, HMSC-200/ton,
SBM-$430/ton, Cottonseed-$300/ton)
Slide compliments – Chad Keifer
2/9/2015
16
© 2014 Regents of the University of Minnesota. All rights reserved.
Barn design
• Cow comfort
• Access to robots
• Minimize lameness
• Labor saving design
Keys to success with robots
© 2014 Regents of the University of Minnesota. All rights reserved.
Free flow system
2/9/2015
17
Free Cow Trafficwith split entry
Automatic
Milking
SystemFeeding
Area
Resting
AreaFetch
Pen
http://www.dairylogix.com
/ http://vetvice.nl/
Fetch cows
Sort
Pen
Slide courtesy D. Kammel
© 2014 Regents of the University of Minnesota. All rights reserved.
2/9/2015
18
Directed Cow Traffic
“Feed First”
Automatic
Milking
System
Selection
Unit
Holding
Pen
Sort
Pen
Resting
Area
Feeding
Area
Slide courtesy D. Kammel
© 2014 Regents of the University of Minnesota. All rights reserved.
Guided flow - Milk first system
Photo courtesy Erica Kiestra, KIE Farms
2/9/2015
19
Directed Cow Traffic
“Milk First”
Automatic
Milking
System
Feeding
Area
Resting
AreaSelection
Unit
Holding
Pen
Sort
Pen
Slide courtesy D. Kammel
© 2014 Regents of the University of Minnesota. All rights reserved.
Free Flow vs Guided Flow
Item Free Flow Guided Flow
Fetch rate (labor) 16%1 8.5%1
Initial
investment
lower higher
Level of mgmtcomplexity
lower higher
Feeding
complexity
higher lower
1Rodenburg, 2007
2/9/2015
20
Other keys to for successCow Comfort
– Overcrowding limits cow movement
– Lameness decreases visits and increases fetch
rates (Bach, 2007) (Borderas 2008)
Excellent ventilation and fly control to minimize
bunching
© 2014 Regents of the University of Minnesota. All rights reserved.
Lameness prevalence
Stall surface Lameness
Severe
Lameness
Sand (14) 22.5%a 4.3%a
Waterbed (7) 35.3%b 10.2%b
Mattress (22) 40.9%b 8.5%b
Mattress + pasture (5) 21.5%a 2.7%a
Bedded pack (3) 19.0%a 1.7%a
2/9/2015
21
Labor savers - drovers lanes & split entry
Labor savers - strategic gates
Compliments David Kammel
2/9/2015
22
© 2014 Regents of the University of Minnesota. All rights reserved.
Labor savers - headlocks
Special needs/treatment pen??
2/9/2015
23
Where should the footbath go?
© 2014 Regents of the University of Minnesota. All rights reserved.
Pre-fresh feeder for heifers
Will this improve training of heifers
• Possibly improve visits of heifers
• Is it worth the cost?
• How much time to you spend training pre-fresh heifers
2/9/2015
24
© 2014 Regents of the University of Minnesota. All rights reserved.
Labor comparison
• AMS dairies (all single box units)
– 485 to 1410 lbs milk per man hour
• Parlor dairies
– >500 cows, 772 to 922 lbs milk per man hour
– 100-300 cows, 551 to 771 lbs milk per man
hour
Caution Preliminary survey results
Unpublished, Piktaranta, 2014
© 2014 Regents of the University of Minnesota. All rights reserved.
Stocking Rate
- In 13 herds with 34 to 71 cows/AMS, higher
stocking densities were associated with lower
milking frequency. (Deming 2013)
- When there are more than
• 60 cows per AMS,
• the number of fetch cows
• Increases (free flow).
• (Rodenburg and Wheeler, 2002)
Fig.1.voluntary milking and cows per box
0
2
4
6
8
10
12
14
16
18
20
45.0 50.0 55.0 60.0 65.0 70.0
cows per milking box
% l
azy
% lazy cows
% lazy milkings
2/9/2015
25
© 2014 Regents of the University of Minnesota. All rights reserved.
Enjoy technology
• Use the information available to
optimize performance and cow
health
Keys to success with robots
© 2014 Regents of the University of Minnesota. All rights reserved.
Focus on maintenance
• Equipment is expensive
• Downtime is more expensive
• Be prepared for higher repair costs
Keys to success with robots
2/9/2015
26
© 2014 Regents of the University of Minnesota. All rights reserved.
Start up & Dealer support
• Use proven tactics at start up
• Successful start up minimizes
financial stress
• Good dealer support is essential
Keys to success with robots
© 2014 Regents of the University of Minnesota. All rights reserved.
Requires excellent business and
management skills
• Higher management than
conventional system
• Mindset to maximize output/robot
Keys to success with robots
2/9/2015
27
© 2014 Regents of the University of Minnesota. All rights reserved.
Mindset Change• Focus on milk per robot per day
• Maximize milk per minute of box time
– Cow milking speed
– Machine settings and maintenance
– Minimize prep time and attach time
• Cow cooperation
• Teat placement and udder balance
• Singe udders
– Minimize percent free time
– Optimize refusals
• Each 5 seconds more/milking = one less cow
© 2014 Regents of the University of Minnesota. All rights reserved.
Milk per AMS
Biggest factors
• Percent idle time
• Milk flow rate per minute of robot use
Balance:
• Milk/cow/visit (milking permission)
• Milkings/cow/day
• Milk/cow/day
• Number of cows/robot
2/9/2015
28
© 2014 Regents of the University of Minnesota. All rights reserved.
Cull RateTypically 0-3% (Rodenburg, 2002)
AMS cannot milk some udder conformations
– Crossed rear teats
– Real deep udder
– Severe reverse tilt
Cull for optimum robot performance
– optimum box time (milk per minute of box time)
– milking speed
– behavior
– attach time
© 2014 Regents of the University of Minnesota. All rights reserved.
Milking time, speed, robot use
Milking time
(minutes)
Milking
speed
(L/min)
Robot idle
time (%)
Average 5.5 2.8 19
Minimum 4.5 2.3 8
Maximum 6.3 4.7 66
Reduce box time by 1 minute/cow increases capacity by 12%
2/9/2015
29
© 2014 Regents of the University of Minnesota. All rights reserved.
Economics
© 2014 Regents of the University of Minnesota. All rights reserved.
Cost/ValueExpensive – compared to what???
Family dairy looking to expand
Trade offs
– labor (hired and family)
– capital investment
– lifestyle
Choices:
– low cost parlor – hired/family labor
– modern parlor – hired/family labor
– AMS – family labor
2/9/2015
30
© 2014 Regents of the University of Minnesota. All rights reserved.
Slide complements of Tranel & Schulte, 2013
© 2014 Regents of the University of Minnesota. All rights reserved.
Inputs – Two robots
• 144 cows
• $200,000/robot
• $21,600 barn cost
• 5.0% opportunity cost
• 10 year’s of loan
payments
• 15 year robot life
• $25,000 residual/robot
• Increase milk 5 lbs/c/d
• Decrease labor 5 hr/d
• Decrease SCC 10%
• $17.50/cwt milk price
2/9/2015
31
© 2014 Regents of the University of Minnesota. All rights reserved.
Outputs
Robotic milking system economic analysisby William Lazarus and Jim Salfer Most recently modified 12/1/2013
Farm Information Dan Dairyman, 11/8/13
Partial budget analysis (before-tax)
Positive Impacts: Negative Impacts:
Increased Incomes: Decreased Incomes Expected: $0
Increased Milk Production $41,580 Increased Expenses:
Increased Milk Premiums $2,737 Increased Repair and Insurance Costs $16,000
Increased Cull Cow Sales (minus = decrease) $0 Change in Feed Quantity Due to DMI Change $10,417
Software Value to Herd Production $5,040 Extra Cost to Pellet the Feed Fed in the Robot Booth $5,227
Total Increased Incomes $49,357 Increased Cow Replacement Costs (minus = decrease) $0
Decreased Expenses: Increased Utilities and Supplies $972
Reduced Heat Detection $2,190 Increased Records Management $3,942
Reduced Labor $27,375 Capital Recovery Cost of Robots (Dep & Int) $38,537
Reduced Labor Management $0 Total Increased Expenses $75,095
Total Decreased Expenses $29,565 Total Negative Impacts $75,095
Total Positive Impacts $78,922 NET ANNUAL FINANCIAL IMPACT w/o Housing = $3,827
Capital Recovery Cost of Housing (Dep & Int) $1,405
Total Capital Recovery Cost of Robots & Housing $40,618
NET ANNUAL FINANCIAL IMPACT with Housing = $1,746
Robot's salvage value at the end of its useful life:
Estimated value at end $50,000 Annualized PV
Estimated value at end, present value (PV) at 5%/year, $24,051 $2,317
NET ANNUAL IMPACT with with Robot's salvage value $4,063
© 2014 Regents of the University of Minnesota. All rights reserved.
Sensitivity AnalysisImpacts of varying key variables over a range of values, showing breakeven points:
-$40,000
-$30,000
-$20,000
-$10,000
$0
$10,000
$20,000
$30,000
$40,000
0 2 4 6 8 10Imp
act
Projected Change in Milk lb/day
NET ANNUAL IMPACTIm
pac
t
NET ANNUAL IMPACT
2/9/2015
32
© 2014 Regents of the University of Minnesota. All rights reserved.
Sensitivity AnalysisProjected Change in Milk lb/day
-$40,000
-$30,000
-$20,000
-$10,000
$0
$10,000
$20,000
6 8 10 12 14 16 18 20
Imp
act
Years of useful life
NET ANNUAL IMPACT
© 2014 Regents of the University of Minnesota. All rights reserved.
Scenario’s
Enter scenario descriptions here=>Increased
investment Less Pel lets
New Barn
($12,000 s ta l l )
No reduced
labor
Lower ini tia l
mi lk and no
more mi lk
Current Last Alternative scenarios
Enter input data for alternative scenarios to compare Data Entries 1 2 3 4 5
Estimated Cost per Robot $200,000 $200,000 $220,000
Related housing changes needed/cow $150 $150 $500 $100 $7,222Increased Insurance Value of Robot & Housing vs.
Current $400,000 $400,000 $440,000 $1,400,000
Anticipated Change in Hours of Milking Labor 5 6 0
Reduced Hours for Labor Management 0 0.6 0.0
Lbs of Milk per Cow per Day, Past Year 75 75 60
Projected Change in Milk Production 5 5 5 16 0
Estimated Percent Change in SCC -10 -5 -15
Pellets Fed in Robot Booth 11 11 6
Total investment for the robots and housing $421,600 $512,000 $414,400 $1,439,968 $421,600 $421,600
NET ANNUAL IMPACT $4,063 $4,426 $16,405 -$17,915 -$23,456 -$18,170
RecalculateReset
2/9/2015
33
© 2014 Regents of the University of Minnesota. All rights reserved.
Summary
• Milking process fits well with robotic technology
• Most current users are satisfied with their decision
– Dairies can expand w/o hiring labor
– Producers can have more flexible schedule
• Whole system approach for best success
• Must make the cash flow work!
• What is the return on investment?
• Requires excellent management!
•
© 2014 Regents of the University of Minnesota. All rights reserved.
Summary
• Adoption rate in U.S. will depend on:
– Availability of labor
– Cost of labor
– Cost of technology
2/9/2015
34
© 2014 Regents of the University of Minnesota. All rights reserved.
Acknowledgements
• Dairy producers
• Lely & DeLaval for assistance with our
project
• GEA & BouMatic
• David Kammel, UW-Madison
• Kelly Froelich
• Lucas Salfer, Tyler Evink, Michael Schmitt,
Andrew Plumski, Nathan Bos
© 2014 Regents of the University of Minnesota. All rights reserved.
Jim Salfer E-mail salfe001@umn.eduPhone: 320.203.6093