field service advisory committee (fsac) meeting · 11.7.1.2 examples of proper recording...
TRANSCRIPT
2012 FSAC Agenda 2.21.2012 1
Field Service Advisory Committee (FSAC) Meeting
March 6, 2012 Synergy I
Embassy Suites Hotel Fort Worth, TX
1. Call to Order – Terry Hopper, Chair, FSAC
2. Agenda Review and Repair
3. Minutes from 2011 FSAC Meeting (attached)
4. QCS Field Service Program Update – Steven Sievert, QCS
5. QCS Meter Center/Technician Program Update – Steven Sievert, QCS
6. Old Business
a. Electronic ID system performance validation b. Other old business
7. New Business
a. Proposed changes in ‘Guidelines for Field Services’ b. Proposed changes in ‘Guidelines for Meter Centers and Technicians’
8. Recess to NADMA Meeting
9. Reconvene FSAC Meeting
10. New Business (continued)
a. Action on proposed changes (if applicable) b. Other new business
11. Adjourn FSAC Meeting
Field Service Advisory Committee (FSAC) Meeting March 15, 2011 Intercontinental Hotel, New Orleans, LA 1. FSAC meeting called to order at 8:20 a.m. by Robert Albrecht, Chair 2. There were no additional agenda items brought from the floor. 3. Steven Sievert, QCS presented minutes from the September 22, 2010 FSAC Meeting. Minutes were
approved as presented. Steven Sievert was appointed to take minutes for the 2011 meeting. 4. Field Services
a. Presentation (attached to minutes) by Steven Sievert, QCS Program Manager & Field Service and Meter Center auditor.
b. Proposed change in Auditing Guidelines for Field Services, page 8 i. ‘For electronic meters checked via the statistical analysis method, this documentation
may be in the form of a computerized spreadsheet, manual listing, or other organized system and must demonstrate that the meters are operating within tolerance.’
c. Discussion on guideline for validation of electronic ID system performance 5. Meter Center and Meter Technicians
a. Presentation (attached to minutes) by Steven Sievert, QCS Program Manager & Field Service and Meter Center Auditor.
6. Committee meeting recessed at 9:35 a.m. 7. Meeting reconvened at 2:15 p.m. 8. Motion to approve recommendation for statistical reports for EMM as presented.
a. Recommendation to include 5% tolerance in the revision. b. M-S-P.
9. Discussion on proposal for EID system validation. a. Steven Sievert presented sample language. b. Motion to have the auditor present proposal to managers during field service audits, gain
input and/or feedback, and present summary at next FSAC meeting. M-S-P. 10. No other new business. 11. Terry Hopper elected as FSAC Chair. 12. Adjourned at 2:55 p.m. Recorded by: Steven Sievert QC Program Manager/Field Service and Meter Center Auditor Quality Certification Services Inc.
2011 March 15 FSAC Minutes 3.21.2011 1
Field Services Auditing Procedures - Version 12.1 Page 3 of 11
Centering Period Months for Field Services
Field service providers are subject to annual audits. Below is a schedule of target months for the audits. Please note the projected on-site audit year listed as even or odd number years. If you have an “Even” next to your affiliate, you would have an on-site audit in 2012, 2014, etc. Like wise for those with “Odd”- 2013, 2015, etc. would be your on-site audit year. Your off-site audit would be held the following year unless notified differently by the auditor.
January .................................................................................................................... (Odd) Minnesota DHIA February ...................................................................................................................... (Odd) Arizona DHIA ................................................... (Odd) Southern DHIA Affiliates - Southeast DHIA and Tennessee DHIA ............................................................................................................................. (Even) Puerto Rico DHIA
March ............................................................................................................................ (Even) Texas DHIA .................................................................................................... (Odd) Integrated Dairy Herd Improvement April ................................................................................................................. (Odd) DHI Cooperative Inc. May ......................................................................................................................... (Even) California DHIA ............................................................................................................................. (Even) San Joaquin DHIA .............................................................................................................................. (Even) Jim Sousa Testing June .......................................................................................... (Odd) NorthStar Cooperative DHI Services .......................................................................................................... (Odd) Indiana State Dairy Association July ................................................................................................................................. (Even) Idaho DHIA August ....................................................................................................................... (Odd) Lancaster DHIA ........................................................................................................................................ (Odd) Tulare DHIA .............................................................................................................................. (Odd) Dairy Lab Services September .............................................................................................. (Even) Dairy One Cooperative Inc. ............................................................................................................... (Even) Gallenberger Dairy Records October ..................................................................................... (Odd) AgSource Cooperative Services/CRI November ................................................................................................... (Even) Washington State DHIA ..................................................................................................................... (Even) Heart of America DHIA .................................................................................................................. (Even) Mid-South Dairy Records ...................................................................................................................... (Even) Rocky Mountain DHIA December .............................................................................................. (Odd) United Federation of DHIAs
Meter Center & Meter Technicians Procedures - Version 12.2 Page 3 of 21
Centering Period Months for Meter Centers – Even Years Meter Centers are subject to mandatory biennial, on-site audits. Below is a schedule of target months for the on-site audits scheduled to occur during even numbered years. January ........................................................................................................................................................... February ...........................................................................................................................Puerto Rico DHIA March ................................................................................................................ Texas DHIA – Stephenville April ...................................................................................................................... Lancaster DHIA Portable May ................................................................................................................................. San Joaquin DHIA June ................................................................................................................................................................ July .............................................................................................................. Mini-Cassia Calibration Center ................................................................................................................. Vanden Bosch Calibration Center August ............................................................................................................................................................ September ......................................................................................................... Dairy One – Ithaca Portable ........................................................................................................................Dairy One – Ithaca Stationary October ........................................................................................................................................................... November ................................................................................................................Heart of America DHIA .............................................................................................................................. Mid-South Dairy Records ..................................................................................................................................Rocky Mountain DHIA ............................................................................................................................................ Tillamook DHIA ................................................................................................................................ Washington State DHIA ........................................................................................................................................... Willamette DHIA December .................................................................................................... Asociación Holstein de México
Meter Center & Meter Technicians Procedures - Version 12.2 Page 4 of 21
Centering Period Months for Meter Centers – Odd Years Meter Centers are subject to mandatory biennial, on-site audits. Below is a schedule of target months for the on-site audits scheduled to occur during odd numbered years. January ............................................................................................................................... Minnesota DHIA February ................................................................................................................................. Arizona DHIA ............................................................................................................................ Southeast DHIA – Portable ........................................................................................................................................... The Udder Tester March ............................................................................................................................................................. April ....................................................................................................... DHI Cooperative Inc. – Stationary ................................................................................................................... DHI Cooperative Inc. – Portable May .................................................................................................... Eastern New Mexico DHIA – Dexter ............................................................................................................ Eastern New Mexico DHIA – Clovis .............................................................................................................. Integrated Dairy Herd Improvement ...................................................................................................................................Texas DHIA – Canyon ....................................................................................................................................Circle H Headquarters June ............................................................................. NorthStar Cooperative Inc. DHI Services – Lansing July ................................................................................................................................................................. August .............................................................................................................Lancaster DHIA – Stationary ........................................................................................................................... California DHIA – Portable ...................................................................................................................................... Kings County DHIA ..................................................................................................................................Central Counties DHIA ............................................................................................................... Southern Counties DHIA – Shafter .................................................................................................................................................. Fresno DHIA .................................................................................................................................................. Tulare DHIA ......................................................................................................................................... Dairy Lab Services September ...................................................................................................................................................... October ......................................................... AgSource Cooperative Services/CRI – Portable - Dorchester .......................... AgSource Cooperative Services/CRI – Mega Test Rig and Calibration Rig - Menomonie November ............................................................... NorthStar Cooperative Inc. DHI Services – Fox Valley December ....................................................................................... United Federation of DHIAs – Portable
11.7.1 Test day policy utilizing proper recording practices when using electronic milk meters and electronic ID simultaneously
11.7.1.1 Definition On test day, utilizing on-farm electronic milk meters with on-farm electronic identification (hereby addressed as ID) information, it has been substantiated numerous times that not all ID programs are complete, accurate and successful. Realizing that there are currently on-farm electronic milk metering devices standards that are in place it is the intent of this publication to substantiate the proper use and guidelines of electronic ID usage on test-day to give the most accurate and precise information possible for use in genetic evaluations and management practices.
11.7.1.2 Examples of proper recording systems/practices on test day
First group of cows entering milking parlor all need to be visually identified and cross-referenced to the electronic ID system; henceforth two stalls are randomly selected to observe on each group to insure proper ID test day procedures – to substantiate verification of visual observation a paper trail or computer notebook protocol would be used to insure accuracy
First group of cows entering milking parlor all need to be visually identified and cross-referenced to the electronic ID system; henceforth every fifth group is visually identified to insure proper ID for test day procedures – to substantiate verification of visual observation a paper trail or computer notebook protocol would be used to insure accuracy
First group of cows entering milking parlor all need to be visually identified and cross-referenced to the electronic ID system; henceforth each first and last animal is visually identified to insure proper ID for test day procedures – to substantiate verification of visual observation a paper trail or computer notebook protocol would be used to insure accuracy
It is advisable that if there are any misidentified animals then proper notification needs to be made to the dairy producer as to the problem discovered and the entire test day needs to be completed using visual identification until the problem is corrected
11.7.1.3 Validation It is advisable the electronic ID system should have inbuilt validation checks/software to ensure each row has the correct cow sequence. Such checks would include but not limited to:
“Cross out” check - in the event cow A is “read” by sensor but withdraws her head and is “passed out” by another cow B, then when cow A enters properly the sequence is corrected….
“Random Check” – the system can be programmed on recording day so that the electronic system selects a % of units at random on each row for checking – operator must verify cow at the selected units (accept button) and only when all selected units are “accepted” is the row allowed out….
“Narrow Entrance funnel” – as most errors occur at entry-gate to row, it is advisable for parlor installations to have entrance funnel of “one cow length” thereby distancing the jostling activity from sensors.
Note – experts from manufacturers should be consulted here to decide and agree on the best way to “build” quality checks into the system.
11.7.2 Test day policy utilizing proper recording practices when obtaining milk samples on individual animals
11.7.2.1 Definition On test day, various sample vial recordings are used in the world-wide marketplace that adhere to proper test day procedures; however as has been identified by various entities, shortcuts are being made that limit proper identification of samples with individual cows on test day. With the milk sample platform being used for disease, genetics, DNA, and health tests along with the routine component test day requirements it is essential that all milk samples collected on test day be properly identified to the corresponding animal that it is collected from using proper collecting procedures. 11.7.2.2 Examples of proper recording of sample vials on test day
Each animal is recorded on sample vial with name or number corresponding to animal ID that is used on-farm that corresponds with proper laboratory practice procedures
Each animal is recorded on sample vial using bar graph information systems that corresponds with proper laboratory practice procedures
Each animal is recorded on sample vial via RFID chip installed or embedded within the sample vial that corresponds with proper laboratory practice procedures
It is advisable that every sample vial is properly identified with the corresponding correct cow ID in whatever system is approved for proper usage that will follow proper laboratory practice procedures
QCS SIEVERT FSAC 03.06.2012
Field Service Advisory Committee March 6, 2012
QCS SIEVERT FSAC 03.06.2012
Components of Field Service Certification
Certified Field Service
Affiliate
Mandatory Annual Audit
Compliance with Code of Ethics &
Uniform Operating
Procedures
Payment of Fees Compliance
with General and Field Service Auditing
Guidelines
2
QCS SIEVERT FSAC 03.06.2012
Majority of the affiliates meet the minimum
Training documentation is dated • No updates to training programs since 2000-2003 – 75% • We need to provide the tools for new technicians to
succeed • Most likely more training than reported to the auditor • QCS recognizes variances between affiliates – just
document what training you provided
What support is needed? • Forms? Presentations? • Online Training Modules?
3
QCS SIEVERT FSAC 03.06.2012
Number of Field Service Affiliate Managers do not attend any organized training
Increases challenges/increases costs of support • Not aware of industry changes • Higher non-compliance issues • Concern in both lab and field service sectors
4
QCS SIEVERT FSAC 03.06.2012
Meters need to be calibrated at least every 12 months
Many affiliates calibrate more often • QCS can handle multiple calibration dates • Use the latest two dates for the interval
Helpful hints
• Don’t forget to record meter center and technician • Don’t forget second calibration checks when required • No alphanumeric numbers if possible • Use the manufacturer’s serial number for QCS whenever
possible
5
QCS SIEVERT FSAC 03.06.2012
Electronic meters need to be calibrated every 12 months • Water Test Calibration • Parlor Report/EMMR/Manufacturer’s Software Report
demonstrating that meters are accurately weighing milk
Growing number of dairy-owned meters used for DHI
Don’t forget to update make, model and number of meters as parlors expand or refurbish
Myths about electronic meters
• Meters will always be in calibration • DRPC will edit out any errors • A 10-day average takes care of all individual cow errors
6
QCS SIEVERT FSAC 03.06.2012
Make Model 2010 2011 Percent
Afikim 2000/9000/Afilite/Fullflow 9,574 11,658 12.9%
Boumatic Perfection/Precision/Smart Control 21,438 29,138 38.3%
Dairymaster Weighall 268 786 0.9% DeLaval MM15/MM25 (VMS, MU480) 8.114 20.800 23.0% GEA Metatron 16,568 24,886 27.5% Germania Accuweigh 66 42 0.1% Lely All Astronaut Models 62 168 0.2% Nedap Level Meter 16 12 <0.1% Surge Dairy Manager 3,046 2,678 3.0% Universal Uniflo/Electric Weigh Meter 388 340 0.4%
Total 56,034 90,554
*not for publication
QCS SIEVERT FSAC 03.06.2012 8
• Approved Models Afiflo 2000 Afiflo 9000 Full Flow (MM 85/MM95) Afilite (Germania’s Essential) Afi 155 & 155i (Sheep & Goat)
• Low-Line Installation Only
• Must Use Afikim Sampler
• There is no meter performance
report available in the current version of the Afifarm software
QCS SIEVERT FSAC 03.06.2012 9
• Approved Models Precision (Model M/M+) Perfection
• Low-Line Installation Only
• Must Use Boumatic Sampler
• Provantage Software offers
meter performance report – limited number of milkings are represented in the estimation of meter deviation
QCS SIEVERT FSAC 03.06.2012 10
• Model under test at ICAR Smart Control (Boumetrix)
• Did not pass 1st ICAR test in
2011
• Boumatic has indicated there will be a redesigned sampler for this meter
• Plans are to retest in 2012
• Most likely will not be able to use the current Perfection 3000 sampler with this meter
QCS SIEVERT FSAC 03.06.2012 11
• Approved Models Weighall
• Both high line and low-line
installations are approved
• No meter performance report in current Dairymaster Milk Manager software
• Calibration is easy with standard weight
• Dairymaster recently changed
the dump cycle in software – affects accuracy of meter
QCS SIEVERT FSAC 03.06.2012 12
• Approved Models Flomaster MM15 (Flomaster Pro) MM25/MM25W/MM27 SG (Sheep & Goat) Delpro MU480
• Must use proper DeLaval
sampler for each model
• All meters are low-line except for Delpro MU480
• There is no meter performance report available from the Alpro system
QCS SIEVERT FSAC 03.06.2012 13
• DeLaval VMS System
• Must use the DeLaval Shuttle for DHI sampling
• Provide a dealer calibration
report annually
• Each robot has 4 DeLaval MM25W meters – each need to be calibrated
QCS SIEVERT FSAC 03.06.2012 14
• Metatron Meter
• Approved Controllers Metatron 12 S21/P21 Dematron 70 Dematron 75
• Low-Line Installation Only
• Must Use GEA/Westfalia
Sampler
• Reports of Sampler Flooding with Higher Milk Flow Rates
• Metatron performance report meets QC requirements
QCS SIEVERT FSAC 03.06.2012 15
• MIone AMS System
• Must use the MIone sampler
• Provide a dealer calibration report annually
• There is no meter performance report available in the current software for the MIone
QCS SIEVERT FSAC 03.06.2012 16
• Approved AMS Models Astronaut/Astronaut A2 Astronaut A3/A3 Next Astronaut A4
• Must be dealer calibrated on an annual basis
• Milk weight measuring uses a weigh jar on a load cell
• Lely Shuttle A for DHI sampling – ICAR approved
Shuttle A Shuttle B
QCS SIEVERT FSAC 03.06.2012 17
• Approved Models Dairy Manager
• Low-Line Installation Only
• No direct meter performance
report in Surge software – must interface with DC305 or PCDart
• Operational manual and calibration instructions are available from QCS
QCS SIEVERT FSAC 03.06.2012
Required for all herds • All test plans are included – even 40’s and 70’s • Data is used for genetic and management research
Good business practice, even for non-processed herds
• Herds may convert from non-processed to processed • Record of herd code assignment • Release and use of records
Common Problems • Missing Agreements – forgot to get the agreement initially • Missing Signatures – especially Field Service Affiliate
signature
18
QCS SIEVERT FSAC 03.06.2012
Monitoring of Electronic ID Systems
• Discussed in 2011 FSAC meeting • Presentation as part of the 2012 NADMA meeting
Four possible options for field services
1. No FS procedure for EID System Validation 2. FS procedure based on ICAR test day policy 3. FS procedure based on ICAR test day policy but
modified to meet needs of individual dairy facilities or cow handling practices.
4. FS Procedure demonstrating validation of accuracy of electronic ID system proposed by field service affiliate and approved by the auditor.
19
QCS SIEVERT FSAC 03.06.2012
A simple, yet vital mission….
Providing a reliable source of information to people interested in the U.S. dairy records industry.
Thank you for your cooperation, support and commitment to the quality certification process!
20
QCS SIEVERT FSAC 03.06.2012 QQQCS QQCSQCQCSQCQCS QCQCS QCS S QCCSQ S CS QC SIEVSISIESIEVSIEVSIEVSIEVSIESI VERTERT ERT ERT ERT ERT ERT FSACFSACFSAFSACFSACFSAC 03.03.03.03.3003 06.206.206.20606.206.260 01201201201210 2
Field Service Advisory Committee March 6, 2012
QCS SIEVERT FSAC 03.06.2012
Components of Meter Center Certification
Certified Meter Center
Mandatory Biennial On-
Site Audit Compliance with Code of Ethics &
Uniform Operating
Procedures
Payment of Fees Compliance
with General and Meter
Center Auditing
Guidelines
2
QCS SIEVERT FSAC 03.06.2012
Service providers are required to notify the auditor of:
Changes in business name, address, phone, email, contacts Changes in authorized personnel – i.e. meter technicians Changes in equipment/instrumentation
Notification within 30 days of change Send changes to QCS Program Manager – Steven Sievert Assures accuracy in billing, website listings, and monitoring instrument performance Allows for cost-efficient scheduling of on-site discretionary audits
3 QCS SIEVERT FSAC 03.06.2012
None proposed or viewed as critical Approved in 2011
Meter center is required to use a manufacturer’s wand or approved closed jar-to-jar system - Implementation on January 1, 2012 New Standard Flow Wand from Waikato
4
QCS SIEVERT FSAC 03.06.2012
Equipment is aging in many meter centers
Vacuum pumps/gauges are failing (4 in 2010, 7 in 2011) Receiver jars showing some age – air leaks, buildup Scales are inaccurate (9 in 2011)
Unapproved meter modification
Modification of parts so the meter samples faster resulting in inaccurate samples. Glues, cement, etc. – weakens the meter and is not approved for Grade A dairies
Cutting corners on meter service Failure to replace annual service kits Failure to replace hoses
5 QCS SIEVERT FSAC 03.06.2012
Meters need to be calibrated at least every 12 months Many affiliates calibrate more often
QCS can handle multiple calibration dates Use the latest two dates for the interval
Helpful hints Don’t forget to record meter center and technician Don’t forget second calibration checks when required No alphanumeric numbers if possible Use the manufacturer’s serial number for QCS whenever possible
6
QCS SIEVERT FSAC 03.06.2012
Model Model 2010 2011 Percent
FOSS Milko-Scope 224 186 0.2% TeSa Milk-O-Meter *Decertified 12/31/10 168 *0 0.0% Tru-Test Auto Sampler (SB & WB Models) 25,114 21,478 23.3% Tru-Test Economy (SB) 2,934 2,048 2.2% Tru-Test Electronic Milk Meter 528 915 1.0% Tru-Test Ezi-Test (SB & WB Models) 8,123 8,224 8.9% Tru-Test Farmer (SB) 6,845 4,517 4.9% Tru-Test Pullout (SB & WB Models) 51,136 46,748 50.7% Waikato MK V 12,013 7,897 8.6% Waikato SpeedSampler 284 276
Total 107,369 92,289
*not for distribution
QCS SIEVERT FSAC 03.06.2012 8
QCS SIEVERT FSAC 03.06.2012
9
2012
Hosted by Holstein Mexico Held in Queretaro, Mexico Bilingual instruction Most likely in early December – coincide with National Holstein show
QCS SIEVERT FSAC 03.06.2012
A simple, yet vital mission….
Providing a reliable source of information to people interested in the U.S. dairy records industry.
Thank you for your cooperation, support and commitment to the quality certification process!
10
QCS SIEVERT FSAC 03.06.2012
Field Service Advisory Committee March 6, 2012
QCS SIEVERT FSAC 03.06.2012
Components of Meter Center Certification
Certified Meter Center
Mandatory Biennial On-
Site Audit Compliance with Code of Ethics &
Uniform Operating
Procedures
Payment of Fees Compliance
with General and Meter
Center Auditing
Guidelines
2
QCS SIEVERT FSAC 03.06.2012
Service providers are required to notify the auditor of:
• Changes in business name, address, phone, email, contacts • Changes in authorized personnel – i.e. meter technicians • Changes in equipment/instrumentation
Notification within 30 days of change
Send changes to QCS Program Manager – Steven Sievert
Assures accuracy in billing, website listings, and monitoring
instrument performance
Allows for cost-efficient scheduling of on-site discretionary audits
3
QCS SIEVERT FSAC 03.06.2012
None proposed or viewed as critical
Approved in 2011 Meter center is required to use a manufacturer’s wand or
approved closed jar-to-jar system - Implementation on January 1, 2012
New Standard Flow Wand from Waikato
4
QCS SIEVERT FSAC 03.06.2012
Equipment is aging in many meter centers
• Vacuum pumps/gauges are failing (4 in 2010, 7 in 2011) • Receiver jars showing some age – air leaks, buildup • Scales are inaccurate (9 in 2011)
Unapproved meter modification • Modification of parts so the meter samples faster resulting
in inaccurate samples. • Glues, cement, etc. – weakens the meter and is not
approved for Grade A dairies
Cutting corners on meter service • Failure to replace annual service kits • Failure to replace hoses
5
QCS SIEVERT FSAC 03.06.2012
Meters need to be calibrated at least every 12 months
Many affiliates calibrate more often • QCS can handle multiple calibration dates • Use the latest two dates for the interval
Helpful hints
• Don’t forget to record meter center and technician • Don’t forget second calibration checks when required • No alphanumeric numbers if possible • Use the manufacturer’s serial number for QCS whenever
possible
6
QCS SIEVERT FSAC 03.06.2012
Model Model 2010 2011 Percent
FOSS Milko-Scope 224 186 0.2% TeSa Milk-O-Meter *Decertified 12/31/10 168 *0 0.0% Tru-Test Auto Sampler (SB & WB Models) 25,114 21,478 23.3% Tru-Test Economy (SB) 2,934 2,048 2.2% Tru-Test Electronic Milk Meter 528 915 1.0% Tru-Test Ezi-Test (SB & WB Models) 8,123 8,224 8.9% Tru-Test Farmer (SB) 6,845 4,517 4.9% Tru-Test Pullout (SB & WB Models) 51,136 46,748 50.7% Waikato MK V 12,013 7,897 8.6% Waikato SpeedSampler 284 276
Total 107,369 92,289
*not for distribution
QCS SIEVERT FSAC 03.06.2012 8
QCS SIEVERT FSAC 03.06.2012
9
2012
Hosted by Holstein Mexico Held in Queretaro, Mexico Bilingual instruction Most likely in early December – coincide with
National Holstein show
QCS SIEVERT FSAC 03.06.2012
A simple, yet vital mission….
Providing a reliable source of information to people interested in the U.S. dairy records industry.
Thank you for your cooperation, support and commitment to the quality certification process!
10
North American DHI Manager’s Association (NADMA) Meeting
March 6, 2012 Synergy 1
Embassy Suites Hotel Fort Worth, TX
12:30 p.m. Lunch Call to Order – Terry Hopper, NADMA Chair Introductions Agenda Repair
Approval of Minutes from the 2011 NADMA Meeting 1:00 p.m. Recess to Field Service Advisory Committee (FSAC) Meeting
Terry Hopper, FSAC Chair and Steven Sievert, National DHIA/QCS 1:30 p.m. Resume NADMA Meeting
Electronic Meters – Use and Interpretation of EMMR Reports Steven Sievert, George Cudoc
2:00 p.m. Validation of Electronic ID Systems – Mark Eisenga, Steven Sievert 2:20 p.m. Field Technician Training Needs – Terry Hopper 2:30 p.m. Health Break 3:00 p.m. Resume NADMA Meeting Dairy Data Alliance Update – Jay Mattison 3:20 p.m. Identification Traceability Update – Jay Mattison 3:30 p.m. Handling of Robotic Records – Terry Hopper, Steven Sievert 3:45 p.m. Data Handling and Transfer Issues – Roundtable with
DRPC Representatives & Moderated by Steven Sievert Questions from Attendees are Welcome/Desired
4:10 p.m. NADMA Elections
Terry Hopper and Scott Taylor have completed their terms Adjourn NADMA Meeting 4:15 p.m. Resume FSAC Meeting
Terry Hopper, Chair and Steven Sievert, National DHIA/QCS 4:30 p.m. Adjourn FSAC Meeting 6:30 p.m. Tru-Test Reception
Minutes of the
North American DHIA Managers Association (NADMA)
March 6th, 2012
Fort Worth, TX Terry Hopper, Chairman of the North American DHIA Managers Association (NADMA) opened this years meeting at 12:50 p.m. after a nice luncheon with a welcome to all in attendance. He introduced the agenda and asked if there were any additions or modifications. Seeing none he went around the room with introductions and a sign-in sheet was circulated. There were no minutes from the previous years meeting to read or approve. Hopper then reminded everyone that the protocol of the meeting is 1 vote per NADMA Member when issues come up for a vote. He then provided the financial update to the members that there has been no income for many years as this is a one time membership organization with no new members. The current balance in the checkbook stands at $1,741.94 and there have been no expenses of the association since 2009. A question came up concerning the use of Facebook to communicate to the membership. This had not been developed beyond setting up the account. The NADMA meeting was recessed in order to convene the Field Services Advisory Committee Meeting. This committee consists of all currently certified QCS Field Service Providers within the DHI system nationwide and provides 1 vote per provider. Hopper was appointed in last years meeting as Chairman of the Field Services Advisory Committee. Steven Sievert was asked to record the minutes of the meeting for later distribution. The NADMA meeting reconvened at 3:15 in order to consider and discuss further any items brought forward from the FSAC meeting. In addition, several presentations were made that have broad interest to the NADMA group. Steven Sievert made a presentation on the use and interpretation of various electronic meter performance reports and the requirements relating to the audit requirements for these farm-owned meters. George Cudoc reported on efforts and success in modifying DC305 reports to provide more informative data as far as meter calibration results in the parlors. Mark Eisenga of NorthStar Cooperative and Michigan DHIA reported on efforts to verify accuracy and performance of electronic ID systems. A discussion ensued regarding difficulties and actual danger issues facing testers in certain parlors in trying to verify eartags against ID systems. Terry Hopper then made a presentation on new training modules that are available on-line to assist managers in initial training as well as certification requirements. The use of these resources and the cooperation of QCS in this effort were appreciated by those present. Jay Mattison then presented an update on the USDA Disease Traceability program. The USDA seems to be back to square one erasing some five years of progress in regards to efforts to report animal movement and ID requirements. This could put the industry at serious risk in the case of a contagious disease outbreak in the U.S.
Mattison shifted gears and reported on efforts of the Dairy Data Working Group (DDWG) and the in regards to ownership and use of dairy records. There is final draft of a cooperative agreement between all industry partners and the group is working on a business plan. This new entity is tentatively referred to as the Dairy Data Alliance (DDA). This is moving rapidly forward in order to ensure contributors to the database of control and ownership of the valuable and important data being collected within dairy industry in the U.S. This also includes possible future health traits and genomic data which will be critical in the future as well as maintaining the confidential nature of such information. The Council on Dairy Cattle Breeding will move forward with plans to house and control the database and then allow USDA computations to be used for cow indexes and sire evaluations to be generated. This will also be used for genomic analysis in the future as this data becomes more and more valuable. Next on the agenda was the introduction by Mattison of Duane Norman as the person who is finishing the work on the new AM/PM factors. Norman recently retired from AIPL and is well known and respected. He mentioned that the 2X factors are done and 3X should be finished with the next 3 weeks or so to begin beta testing and comparison work. The last item on the agenda was election of officers. Hopper reviewed NADMA officers up for re-election. Bruce Dokkebakken made a motion that given the close function of NADMA and the Field Services Advisory Committee that NADMA be suspended and that the FSAC becomes the platform for future meetings and discussion relating to all NADMA and FSAC concerns. The motion was seconded by Susan Lee, and an individual vote of all NADMA members present was taken with a unanimous vote given to suspend. NADMA will be in suspension until such time as needed or until such time as it is formally disbanded. Terry Hopper’s term as FSAC Chairman expires in 2013 and he will up for reelection at the next FSAC Meeting. He will coordinate with Steven Sievert concerning future meetings and agendas. Chairman Hopper adjourned the meeting at 4:15 p.m. Respectfully Submitted, Scott Taylor, Secretary Note: Steven Sievert, QCS Program Manager will be posting the NADMA meeting minutes and copies of all presentations from the NADMA meeting on the QCS website in conjunction with the FSAC meeting materials.
QCS SIEVERT FSAC 03.06.2012
North American DHI Manager’s Meeting March 6, 2012
QCS SIEVERT FSAC 03.06.2012
Make Model 2010 2011 Percent
Afikim 2000/9000/Afilite/Fullflow 9,574 11,658 12.9%
Boumatic Perfection/Precision/Smart Control 21,438 29,138 38.3%
Dairymaster Weighall 268 786 0.9% DeLaval MM15/MM25 (VMS, MU480) 8.114 20.800 23.0% GEA Metatron 16,568 24,886 27.5% Germania Accuweigh 66 42 0.1% Lely All Astronaut Models 62 168 0.2% Nedap Level Meter 16 12 <0.1% Surge Dairy Manager 3,046 2,678 3.0% Universal Uniflo/Electric Weigh Meter 388 340 0.4%
Total 56,034 90,554
*not for publication
QCS SIEVERT FSAC 03.06.2012
Annual Water Test Calibration by Dealer (54%) Statistical Monitoring through Computerized Report (46%)
◦ Manufacturer’s Software (Dairyplan, Provantage)
◦ Third Party Software with Interface to Manufacturer’s
Software
QCS SIEVERT FSAC 03.06.2012
Electronic meters need to be subjected to a routine calibration check on an annual basis (at least once every 365 days)
• Water Test Calibration • Parlor Report/EMMR/Manufacturer’s Software Report demonstrating
that all meters are operating within 5% tolerance
Records generated using electronic meters exceeding 14 months between calibration checks must be noted as non-certified.
Don’t forget to update make, model and number of meters as parlors expand or refurbish
Myths about electronic meters
• Meters will always be in calibration • DRPC will edit out any errors • A 10-day average takes care of all individual cow errors
4
QCS SIEVERT FSAC 03.06.2012
Date Herd name or Herd code Animal ID Stall or meter ID Measured milk weight Number of milkings represented at each stall/meter Deviation for each stall/meter
Optional ◦ Defined tolerance ◦ ID errors (missing cows, duplicate reads, wrong pens) ◦ Reattachment and manual detach incidents ◦ Milking time deviations ◦ Milking speed ◦ Cross reference with milk shipped weights integrated into
the report or software program
QCS SIEVERT FSAC 03.06.2012
The in-place milk meter is only part of a linked system that includes…
Calibrated Milk Meter
Functioning Milk Meter Controller
Software and Interfaces
Accurate ID
System
Milker (Human) Performance
QCS SIEVERT FSAC 03.06.2012
Expected Milk Weight (MW) this milking Cow MW Ave for Last 10 milkings * (Herd Ave Milk this
milking/Herd Ave for Last 10 milkings) Deviation from Expected Individual MW – Expected MW Meter Deviation (%) Sum (Expected Deviations for this meter)/(Sum Expected MW
for this meter) *100
QCS SIEVERT FSAC 03.06.2012
Usually two methods of ID systems
Electronic ID systems
◦ Manufacturer ID – transponders ◦ Third Party RFID tags and readers ◦ Primary Source of Error – TECHNOLOGY
Manual ID entry
◦ Cow ID is keyed on the controller in the milking stall ◦ Usually leg bands or visual cow number ◦ Primary Source of Error - HUMAN
QCS SIEVERT FSAC 03.06.2012
23 out 24 Animals Read Correctly
QCS SIEVERT FSAC 03.06.2012
GEA Westfalia - Metatron ◦ Dairy Plan
Boumatic - Perfection 3000 ◦ ProVantage
Resources are available from QCS on obtaining and
interpreting these reports
QCS SIEVERT FSAC 03.06.2012
QCS SIEVERT FSAC 03.06.2012
QCS SIEVERT FSAC 03.06.2012
Interface with manufacturer’s software
◦ GEA Westfalia (Dairy Plan) ◦ Afikim (Afimilk, Afifarm) ◦ Boumatic (2045 Agri-Comp, 2050 Provantage, Metrix) ◦ DeLaval (Flomaster, Alpro) ◦ DairyMaster (Milk Manager) ◦ Surge (InFARMation) ◦ Universal (QuadTrac)
Short list of software vendors at present
◦ Dairy Comp 305 (Valley Ag Software) ◦ PCDart (Dairy Records Management Systems)
QCS SIEVERT FSAC 03.06.2012
QCS SIEVERT FSAC 03.06.2012
QCS SIEVERT FSAC 03.06.2012
It does not mean the meter is out of calibration…
◦ But if one meter is out of tolerance on the report, the whole report is not usable for QC program
Time to be a detective and ask some questions…
◦ Primary contacts Dairy Manager Representative(s) from Herd Recording Organization
◦ Secondary contacts Milker(s) Representative from Equipment Manufacturer
QCS SIEVERT FSAC 03.06.2012
Robotic Herds
◦ QCS staff is serving on a ICAR Task Force to develop a computerized solution for AMS Herds
Herds with incomplete identification or EID system challenges Herds with one or more failing/non-communicating controllers Herds with one or more missing or out-of-service meters Herds where all strings do not have the same milking
frequency (some strings 4x, some strings 2x) Parlors where two or more herds are comingled but the report
generated is on a herd code basis (PCDart)
QCS SIEVERT FSAC 03.06.2012
Inaccurate ID reads from automated system
Incomplete herd ID Duplicate animal ID Data entry errors by
milking personnel
QCS SIEVERT FSAC 03.06.2012
Reattachment of milkers – Is the total milk weight listed?
Treated cows – do they bypass the meter?
Incomplete letdown by cows
QCS SIEVERT FSAC 03.06.2012
Milkers not trained in use of equipment
Relief milkers Data entry errors Missing milkings
QCS SIEVERT FSAC 03.06.2012
Meters installed properly?
Meter blockage? Meter out of
calibration? Modifications to
milking system? ◦ Vented inflations? ◦ Milk line size (1” vs.
3/4” vs. 5/8”)? ◦ Settings for
automatic take-offs (milk-flow rate to trigger cluster removal)?
QCS SIEVERT FSAC 03.06.2012
Upload/interface errors ◦ ID data not transferred
properly ◦ Milk weights not
transferred ◦ Stall identification errors
Software upgrades or modifications
QCS SIEVERT FSAC 03.06.2012
Advantages Disadvantages
Low cost Frequency – i.e. monthly Ease of producer access Ease of DHI personnel
access Targeted repair costs Identify weaknesses in the
entire linked system Service opportunity for herd
recording organization – build values in DHI program
Does not clearly indicate whether a meter is operating within tolerances ◦ Part of the process ◦ Not the answer or result
There is no meter system certification or validation without… ◦ communication ◦ interpretation ◦ action ◦ follow-up
QCS SIEVERT FSAC 03.06.2012
A simple, yet vital mission….
Providing a reliable source of information to people interested in the U.S. dairy records industry.
• Performance & Quality Standards
• Compliance Auditing of Providers
• Education, Training, & Development
EMMR Reports
NADMA Fort Worth, Texas March 6, 2012
EMMR
Circumstances where normal statistical calibration report creation is inaccurate yet the meters are likely in calibration Multiple parlors with single data cow files Multiple milking schedules within the herd May be others
Solution
Customize the reports to reflect the nature of the milking operation Can be done with DC305 for sure PcDart capabilities have been unexplored Manufacturer's reports have not been attempted
Scenario #1
A single herd file with cows milked in 2 parlors PARLOR\WM3V Normal reporting for the third milking We can separate the reporting by what actually happens in each parlor thus adding accuracy. PARLOR\WM3V\I2 PARLOR\WM3V\I1
Scenario #2
Different milking schedules for one session to another (different milking frequency not part of the equation as it is a milking schedule issue)
Herd milks 2 sessions each day Pens 2, 3, 6, 7, milk during 1st session Pen 2 milk twice during 1st session
Pens 1, 3, 4, 6, 7 milked during 2nd session Pens 1, 4 milked twice during 2nd session
#2 Solution
PARLOR\WM1V changed to: PARLOR\WM1V FOR (PEN=2)(PEN=3)(PEN=6)(PEN=7)
PARLOR\WM2V changed to: PARLOR\WM2V FOR (PEN=1)(PEN=3)(PEN=4)(PEN=6)(PEN=7)
Electronic ID Verification
What Now?
• 95% is not good enough • Is it safe in the parlor for the tester to get numbers? • Will you slow up production? • Is it time to abandon the test?
NorthStar Cooperative Policy • Goal is to provide the most accurate data
available to the farmer.
• When in the field we ask and expect our Information Specialist to fill out The Electronic ID form for all herds that have AMR on the farm.
• Collect the PC-Dart 817 report, the Dairy Comp 305 parlor deviation report, or the Dairy Plan calibration report, or any other version that is on farm.
NorthStar Cooperative Policy • If problems come up we use these reports
to present to the farmer of proof that Electronic Meters and/or ID system are not working properly.
• We have tried to set the bar high by doing this every test day and not just once or twice a year.
Northstar DHI Services Electronic ID Verification
Herd Owner: ___________________ Herd Number: ___________
Technician : ___________________ Technician Number: ______
Date of Test: ___________________
Stall # Elect. ID Visual ID Correct Stall # Elect. ID Visual ID Correct 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 Correct Identification ________ out of ________ stalls % Correct __________ G:\Customer service\DHI\QCS\Electronic ID Verification
About QCS
Training & ResourcesField Services
LaboratoriesMeter Centers & Technicians
ELISAProcessing Centers
Approved Devices
Meter Tech Training
Auditing Guidelines
Certified Providers
Samples Unknown
ELISA Proficiency
Advisory Committees
Contact Us
Dairy Links
DHI Field and Lab Technicians that have completed training for Johne's (MAP) ELISA Testing
-----------------------------------------------
Johne's Disease Basics Training Module
-----------------------------------------------
Johne's ELISA (MAP) DHI Field Technician Training Module
-----------------------------------------------
Johnes's ELISA (MAP) DHI Laboratory Technician Training Module
____________________________________________________________
About QCS
Training & Resources
Approved Devices
Meter Tech Training
Auditing Guidelines
Certified ProvidersField Services
LaboratoriesMeter Centers
Meter TechniciansProcessing Centers
ELISA TestingJohne's Training
Samples Unknown
ELISA Proficiency
Advisory Committees
Contact Us
Dairy Links
Data Current as of 2/29/2012
Field Service Name Association City State or Province
Johne's Disease Basics
Johne's DHI Field
Tech
Johne's ELISA
Lab Tech
Minnesota DHIA Todd Beaver Barron-Washburn DHIC Barron WI 1/1/2012
DeAnn Nelson Barron-Washburn DHIC Barron WI 12/1/2011
Karen Riggle Becker County DHIA Ponnsford MN 2/29/2012
Charles Clobes Fillmore County DHIA Lewiston MN 11/29/2011 1/12/2012
Valerie Wendt Fillmore County DHIA Spring Valley MN 1/13/2012
Charles Sell Houston County DHIA Brownsville MN 1/30/2012 1/30/2012
Amy Sylling Houston County DHIA Spring Grove MN 2/8/2012 2/8/2012
Gene Stoeckel Isanti County DHIA Princeton MN 1/18/2012 1/18/2012
Kathy Stoeckel Isanti County DHIA Princeton MN 1/21/2012 1/21/2012
Angie Barlau Minnesota DHIA Buffalo MN 2/14/2012 2/14/2012
Harley Bork Minnesota DHIA Buffalo MN 12/15/2011 12/15/2011
Bruce Dokkebakken Minnesota DHIA Buffalo MN 2/20/2012 2/18/2012
Rosalie Egge Minnesota DHIA Cannon Falls MN 11/27/2011
Greg Freiburg Minnesota DHIA Buffalo MN 2/21/2012 2/21/2012 2/20/2012
Gabe Gieske Minnesota DHIA Buffalo MN 12/15/2011
Ron Gruber Minnesota DHIA Buffalo MN 2/27/2012
Eve Hennen Minnesota DHIA Montgomery MN 1/31/2012 1/31/2012
Ernie Hikkula Minnesota DHIA Menahga MN 2/20/2012 2/20/2012
Anna Holicky Minnesota DHIA Le Center MN 2/28/2012 2/28/2012
Page 1 of 2
Update on Animal Disease Traceability March 6, 2012 Prepared by NDHIA - QCS Staff
The cart is loaded but……
Disease Traceability
APHIS has moved to ADT Heavy on “distance” State/Tribe Animal Health Partners Administer of tags Different by tag
Happy to be a helper USAHA ID committee
Summary
Help meet the objective: Assuring that valid and single IDs for dairy cattle
are available and data flow as appropriate
Update on AM/PM Factors
March 6, 2012 Prepared by NDHIA - QCS Staff AIPL Staff and Mike Schutz
Progress Report on AM-PM Factors Mike Schutz Purdue University Over the past year, work has been completed on the comparison of DeLorenzo and Wiggans ratio factors with the Liu factors (details about methods in Table 5). An abstract of that work is presented below. The chief conclusions of that work were that:
Daily milk yields predicted from AM and PM milkings by the DeLorenzo/Wiggans ratio methods and Liu Method consistently had higher correlations with true daily yields, smaller absolute errors and smaller rMSE (measures of goodness of fit) for first and second or later parity compared to current factors when applied to the data from which they were derived. Of the ratio factors, DeLorenzo Method 1 and 2, Method 1 consistently led to more accurate prediction of daily milk yield than the alternative Method 2. Compared to DeLorenzo and Wiggans Method 1, the Liu method gave higher correlations with true daily yields, smaller SD of estimates, smaller absolute errors, and smaller rMSE for the file from which the factors were derived, but the advantages were diminished when applied to the other half of the data. Because the Liu Method depends on coefficients of regression within 168 ParityxMilking IntervalxDIM subclasses, smoothing across these subclasses to avoid large differences in adjacent subclasses within parity can be problematic, especially at Milking interval or DIM extremes where observations are limited. Further, the implementation of the Liu factors would be more difficult. Thus implementation of DeLorenzo/Wiggans Method 1 factors by parity will be more straight forward. These are essentially derived the same way as the factors currently in use. The major difference in the new factors, besides that they are derived from individual cows instead of herd average milking times, is that they will be separate for Parity 1 and Parity 2+.
Preliminary milk factors are shown in Tables 2 and 3 for milk yield for AM and PM, respectively. Tables 4 through 6 show the results of fitting the current and estimated factors on the data from which they were derived and a separate subset of the data for testing. The final step for 2x factors is to smooth the Days Open Adjustments. The 2x Factors for fat and protein will be finalized in a similar way for the subset of data for which component values were obtained. The files of factors will be sent to Duane Norman as files with factor values in machine language:
1. Milking interval factors for AM milking (parity, interval class, factor) 2. Days open factors for AM milking (parity, interval class, slope) 3. Milking interval factors for AM milking (parity, interval class, factor) 4. Days open factors for AM milking (parity, interval class, slope)
Factors for 2x milk, fat, and protein will be sent to Dr. Norman by March 6, 2012. Factors for 3xmilk, fat, and protein will be sent to Dr. Norman by April 1, 2012.
M.M. Schutz*1 and H.D. Norman2; Purdue University, West Lafayette, IN, USA1, USDA-ARS, Beltsville, MD, USA2
Verification of factors to estimate daily milk yield from one milking of cows milked twice daily The objective of this research was to verify factors to predict daily milk yield when milk is sampled once per d for cows milked twice (2x) per d. Milk weights for both milkings were recorded automatically by 30 herds and collected by Dairy Herd Improvement supervisors. Data was split into 2 subsets for developing (FACT) and testing (TEST) factors. Following edits, 179,064 daily milk weight records of 2941 first lactation (L1) cows and 298,905 records of 4757 later lactation (L2) cows remained in FACT and 177,299 records 2120 L1 cows and 335,692 records of 3319 L2 cows remained in TEST. Factors currently in use to adjust single milking yields for milking interval (MINT) were applied. Also, 3 methods were compared to estimate factors or equations to predict daily milk yield. First, factors were estimated as the ratio of the sum of daily yield to the sum of partial yield within a parity-MINT class (13 intervals in 2 parities) [Method 1] or as the sum of the ratios of daily yield to partial daily yield for each cow-day divided by the number of cow-days within parity-MINT class [Method 2]. Resulting factors from both methods were smoothed, applied to data, and residuals were regressed on days in milk (DIM) for FACT and applied to TEST. Regression equations (n=168) were also developed within parity-MINT-DIM classes (2x7x12) [Method 3] to jointly account for MINT and DIM. Separate factors were derived for milking 1, and 2, for L1 and L2. Method 3 resulted in consistently strongest correlations between estimated and actual yields, and smallest variances of estimates, and root mean squared errors (rMSE) for milkings in L1 and L2 for FACT. When applied to TEST, Method 1 resulted in rMSE of 2.07 (Milking 1, L1), 2.12 (Milking 2, L1), 2.64 (Milking 1, L2), and 2.85 kg (Milking 2, L2); compared to rMSE of 2.13, 2.26, 2.68 and 2.83 kg, respectively, from current factors for the same milkings for L1 and L2. Methods 1 and 3 provide more accurate prediction of daily milk weight from a single milking for herds milking 2x daily than factors currently in use. KEYWORDS milking interval adjustment factor milking frequency
Table 1. Distribution of numbers of cow-test-days with intervals prior to AM or PM milking for all breeds.
Factor File Test File Milking Interval
Class (hr.)
AM
PM
AM
PM
<9.0 420 133 9.0-9.24 325 110
9.25-9.49 662 3811 364 3258 9.5-9.74 1186 462 1010 2342
9.75-9.99 1785 790 2047 4024 10.0-10.24 2517 2034 4324 6341
10.25-10.49 3738 4272 8641 8734 10.5-10.74 5240 10966 15619 12091
10.75-10.99 8363 24002 24904 21250 11.00-11.24 14543 42930 37680 44697 11.25-11.49 26170 61756 57550 75856
11.5-11.74 41153 77851 53506 58974 11.75-11.99 63844 78771 32424 35879 12.00-12.24 79507 61825 35431 34158 12.25-12.49 77423 42664 61294 53038
12.5-12.74 60991 26326 74646 58102 12.75-12.99 42569 14295 44759 36371
13.0-13.14 23977 8367 21469 24106 13.25-13.49 11106 5400 12367 15887
13.5-13.74 4469 3544 8802 8991 13.75-13.99 2123 2699 6266 4814
14.0-14.24 1383 2042 3948 2036 14.25-14.49 574 1239 2316 953
14.5-14.74 3901 776 3381 370 14.75-14.99 447 168
>15 700 551 Total 477969 477969 512991 512991
Table 2. Current milking interval factors and preliminary (D-W method b) factors by parity and breed for AM milking.
Factor file Milking Interval
Class (hr.)
Current Parity 1 Parity >=2
<9.0 2.465 9.0-9.24 2.465
9.25-9.49 2.465 2.2876 2.3036 9.5-9.74 2.411 2.2519 2.2667
9.75-9.99 2.359 2.2172 2.2311 10.0-10.24 2.310 2.1836 2.1965
10.25-10.49 2.262 2.1510 2.1630 10.5-10.74 2.217 2.1194 2.1305
10.75-10.99 2.173 2.0887 2.0990 11.00-11.24 2.131 2.0588 2.0683 11.25-11.49 2.091 2.0298 2.0386
11.5-11.74 2.052 2.0016 2.0097 11.75-11.99 2.014 1.9742 1.9816 12.00-12.24 1.978 1.9475 1.9543 12.25-12.49 1.943 1.9215 1.9277
12.5-12.74 1.910 1.8962 1.9019 12.75-12.99 1.877 1.8716 1.8767
13.0-13.14 1.846 1.8476 1.8522 13.25-13.49 1.815 1.8242 1.8283
13.5-13.74 1.786 1.8014 1.8050 13.75-13.99 1.757 1.7792 1.7823
14.0-14.24 1.730 1.7575 1.7602 14.25-14.49 1.703 1.7363 1.7386
14.5-14.74 1.677 1.7156 1.7176 14.75-14.99 1.652 1.6954 1.6970
>15 1.628 1.6757 1.6769
Table 3. Current milking interval factors and preliminary (DW method b) factors by parity and breed for PM milking.
Factor File Milking Interval
Class (hr.)
Current Parity 1 Parity >=2
<9.0 2.594 2.4855 2.4851 9.0-9.24 2.534 2.4431 2.4419
9.25-9.49 2.477 2.4021 2.4001 9.5-9.74 2.423 2.3625 2.3598
9.75-9.99 2.370 2.3241 2.3207 10.0-10.24 2.321 2.2870 2.2830
10.25-10.49 2.273 2.2510 2.2464 10.5-10.74 2.227 2.2162 2.2110
10.75-10.99 2.183 2.1824 2.1768 11.00-11.24 2.140 2.1496 2.1435 11.25-11.49 2.099 2.1178 2.1113
11.5-11.74 2.060 2.0870 2.0800 11.75-11.99 2.022 2.0570 2.0496 12.00-12.24 1.986 2.0278 2.0201 12.25-12.49 1.951 1.9995 1.9914
12.5-12.74 1.917 1.9720 1.9636 12.75-12.99 1.884 1.9452 1.9365
13.0-13.14 1.852 1.9191 1.9101 13.25-13.49 1.822 1.8937 1.8845
13.5-13.74 1.792 1.8690 1.8595 13.75-13.99 1.763 1.8449 1.8352
14.0-14.24 1.736 1.8214 1.8115 14.25-14.49 1.709 1.7985 1.7884
14.5-14.74 1.683 1.7762 1.7659 14.75-14.99 1.683
>15 1.683
Table 4. Correlation of estimated daily milk from a single milking with actual daily yield, standard deviation of yield estimates, and square root of Mean Squared Error from application of 4 different sets of factors to estimate daily milk from a single AM milking.
Factor File Test File rtrue,estimated SD of
estimates Root MSE
rtrue,estimated SD of estimates
Root MSE
Parity 1 Current Factors .966 18.22 4.74 .958 16.20 4.70 D-W Method a .967 18.16 4.61 .958 15.91 4.57 D-W Method b .967 18.28 4.67 .958 16.02 4.64
Liu Method .967 16.91 4.45 .957 14.85 4.46 Parity >= 2
Current Factors .972 25.35 5.99 .972 24.88 5.91 D-W Method a .972 25.33 5.93 .972 24.58 5.81 D-W Method b .972 25.53 6.04 .972 24.78 5.92
Liu Method .973 23.80 5.67 .972 23.11 5.72 Table 5. Correlation of estimated daily milk from a single milking with actual daily yield, standard deviation of yield estimates, and square root of Mean Squared Error from application of 4 different sets of factors to estimate daily milk from a single PM milking.
Factor File Test File rtrue,estimated SD of
estimates Root MSE
rtrue,estimated SD of estimates
Root MSE
Parity 1 Current Factors .957 17.88 5.22 .950 15.70 4.98 D-W Method a .958 17.99 5.14 .951 16.03 4.97 D-W Method b .958 18.10 5.19 .951 16.12 4.99
Liu Method .959 16.78 4.93 .949 14.89 4.84 Parity >=2
Current Factors .963 24.98 6.71 .967 24.36 6.24 D-W Method a .964 25.06 6.66 .967 24.70 6.28 D-W Method b .964 25.22 6.74 .967 24.84 6.33
Liu Method .965 23.62 6.39 .966 23.31 6.22 Current Factors: Factors taken from AIPL website, in current use, and derived from DeLorenzo and Wiggans (1985). The same set of factors is used for all parities. D-W Method a: Factors calculated as Factor = Σy/Σx for each interval by parity subclass, where x is individual milking weight and y is daily weight. Factors were smooth by weighted regression of their reciprocals on interval, as in DeLorenzo and Wiggans. DIM was regressed on residuals to develop stage of lactation adjustment (not shown). There were 24 milking interval classes and 2 parities. D-W Method b: Factors calculated as Factor = Σ(y/x) / n for each interval by parity subclass, where x is individual milking weight and y is daily weight. Factors were smooth by weighted regression of their reciprocals on interval, as in DeLorenzo and Wiggans. DIM was regressed on residuals to develop stage of lactation adjustment (not shown). There were 24 milking interval classes and 2 parities. Liu method: y = b0 + b1x, where y is actual daily yield and x is partial yield and a slope and intercept are kept (not smoothed) for each parity by milking interval by DIM subclass. For this model there were 2 parities, 7 milking interval classes, and 12 DIM subclasses. [Note: Number of DIM subclasses may need to be reduced to match records for fat and protein that had far fewer records.]
Table 6. Absolute prediction errors and absolute difference from yield estimated with Current Factors for herds milked 2x for all breeds using a test file and factor file. Factor File Test File AM PM AM PM Abs error Abs diff
from Current
Abs error
Abs diff from
Current
Abs error Abs diff from
Current
Abs error
Abs diff from
Current Parity 1
Current Factors 3.29 NA 3.58 NA 3.34 NA 3.57 NA D-W Method a 3.17 0.82 3.50 0.89 3.26 0.93 3.56 0.99 D-W Method b 3.24 0.68 3.53 1.16 3.33 0.80 3.57 1.20
Liu Method 3.13 1.43 3.46 1.47 3.29 1.48 3.54 1.57 Parity >= 2
Current Factors 4.21 NA 4.67 NA 4.10 NA 4.38 NA D-W Method a 4.16 0.59 4.64 0.68 4.07 0.89 4.42 0.93 D-W Method b 4.29 0.76 4.68 1.01 4.17 0.83 4.45 1.19
Liu Method 4.06 1.64 4.60 1.69 4.19 1.90 4.53 1.85
Table 7. Correlation of estimated daily milk from a single milking with actual daily yield, standard deviation of yield estimates, and square root of Mean Squared Error from application of 4 different sets of factors to estimate daily milk from a single AM milking.
Factor file Test File
Mean SD Mean SD Parity 1 AM weight (lbs)
35.63
9.73
33.56
8.24
PM weight (lbs) 31.95 8.65 30.19 8.13 Daily weight (lbs)
67.58 17.50 63.75 15.30
Interval to AM (h)
12.6 .804 12.6 .833
Interval to PM (h)
11.4 .805 11.4 .841
Days in Milk
188.61 101.34 175.73 98.7
Parity 2 AM weight (lbs)
39.13
13.5
35.10
12.75
PM weight (lbs) 34.75 11.93 31.92 12.28 Daily weight (lbs)
73.89 24.46 67.02 23.94
Interval to AM (h)
12.75 .720 12.56 0.914
Interval to PM (h)
11.25 .731 11.44 0.921
Days in Milk Percentage Holstein Jersey Brown Swiss
185.40
89.9 8.32 0.65
103.50 184.97
65.7 32.1 0.82
102.84
[Most Jerseys from 1 large herd so difficult to balance across files when splitting by herd].
Update on Dairy Data Alliance and Cooperative Agreement March 6, 2012 Prepared by NDHIA - QCS Staff
Background
Council on Dairy Cattle Breeding (CDCB) Implemented to fill the role of industry
consensus group in the 1980’s Members include: PDCA (breeds) NAAB (AI organizations) DHI (Milk recoding)
DRPCs (observers) USDA – AIPL (observers)
Background
Council on Dairy Cattle Breeding (CDCB) Became more formal on consensus in the
1990s Quality certification of dairy records
become a function of the CDCB in 1998 CDCB sub-contracts with QCS to do the
quality certification Provides data flow to USDA-AIPL
Data Flow Today
MOU between AIPL and the Council Council provides quality certified data AIPL maintains the database AIPL calculates genetic evaluations AIPL makes genetic evaluations available
to the public AIPL provides on-going research
CDCB formed a Dairy Data Working Group
1. Look at the current system of data flow
2. Determine options for sustainable dairy
data system in the future
Challenges to the Current System Database Data contributors need to control how their data
are used Genomics Non-Council members want to directly submit
data to AIPL Non-data contributors benefit without making
contributions AIPL funding for the future 2013 Genomics agreement expires Status Quo is NOT an option
DDWG Recommendations to the Council
1. Approve a new Cooperative Agreement with AIPL.
2. Approve the formation of the CDCB structure to support the US dairy industry.
Cooperative Agreement Council members own the database Council supplies staff collaborating with AIPL to support
calculation of genetic evaluations Council through the Alliance distributes genetic
evaluations Genetic and genomic evaluations are distributed by the
Alliance Helps maintain funding for AIPL, by letting AIPL focus on
research Term – five years, then renewable
Dairy Data Alliance - Objectives
Provide a structure for world leading management tools and genetic evaluations.
Long-term availability of a national database
Protect sensitive data Recognize value of data and services
Dairy Data Flow Structure – Current
Dairy Data Flow Structure – Recommended
Dairy Data Alliance - Funding CDCB Business Plan Working Group Evaluating and making options Each CDCB Alliance Board representative
group is responsible for percentage of capitalization
Each representative group decides how to apportion capitalization among its members
Dairy Data Alliance – Timetable
First Six Months Council approval CA signed Incorporated and
Operating Agreement Board of Directors Staff Fee schedule Terms for customers
First Six Months (cont…)
Plan to equitably
handle data within the Alliance
Plan to include additional members
Summary
1. Cooperative Agreement implemented 2. Establish a sustainable data flow for the US
dairy industry as part of the CDCB mission. This may include forming the Dairy Data Alliance as part of the CDCB.
Summary
Help meet the objective: Assuring that high quality genetic evaluations
and management tools for the U.S. dairy industry will be available in the future.