Responsible robotics design-A systems approach to developing design guides for robotics in pasture-grazed dairy farming.

Application of robotics and automation in pasture-grazed agriculture is in an emergent phase. Technology developers face significant challenges due to aspects such as the complex and dynamic nature of biological systems, relative cost of technology versus farm labor costs, and specific market characteristics in agriculture. Overlaying this are socio-ethical issues around technology development, and aspects of responsible research and innovation. There are numerous examples of technology being developed but not adopted in pasture-grazed farming, despite the potential benefits to farmers and/or society, highlighting a disconnect in the innovation system. In this perspective paper, we propose a "responsibility by design" approach to robotics and automation innovation, using development of batch robotic milking in pasture-grazed dairy farming as a case study. The framework we develop is used to highlight the wider considerations that technology developers and policy makers need to consider when envisaging future innovation trajectories for robotics in smart farming. These considerations include the impact on work design, worker well-being and safety, changes to farming systems, and the influences of market and regulatory constraints.

Analysis of adoption trends of in-parlor technologies over a 10-year period for labor saving and data capture on pasture-based dairy farms.

The use of precision technology is increasingly seen as an option to improve productivity, animal welfare, resource use efficiency, and workplace features on dairy farms. There is limited research related to longitudinal adoption patterns of precision dairy technologies and reasons for any patterns. The aim of this analysis was to investigate trends in technology adoption regarding both the amount (number of farms with a technology) and intensity (number of technologies per farm) of adoption. Surveys of parlor technology adoption were conducted on New Zealand dairy farms in 2008, 2013, and 2018, with 532, 500, and 500 respondents, respectively. Technologies were grouped into labor-saving (LS, such as automatic cluster removers) or data-capture (DC, such as in-line milk meters) categories. Trends were examined for farms that had only LS, only DC, or LS+DC technologies. Technology adoption increased over time; the likelihood of technology adoption in 2018 (and 2013 in parentheses) increased by 21 (22), 7 (68), and 378% (165) for LS, DC, and LS+DC technology groups, respectively, compared to 2008. Farms with LS+DC technologies also had a greater proportion of LS technologies compared to non-LS+DC farms, although this relationship declined over the 10-yr period. The use of a rotary versus herringbone parlor was estimated to be associated with 356 and 470% increase in the likelihood of adopting LS technologies and LS+DC, respectively, from 2008 to 2018. Regional differences in adoption were also found, with the likelihood of adopting DC and LS+DC technologies found to be 46 and 59% greater, respectively, in the South Island of New Zealand, compared to the base region of Waikato. The results highlight the importance of understanding spatial and temporal farm characteristics when considering future effect and adoption of precision dairy technologies. For example, the analysis indicates the occurrence of 2 trajectories to technology investment on farms, where larger farms are able to take advantage of technology opportunities, but smaller farms may be constrained by factors such as lack of economies of scale, limited capital to invest, and inability to retrofit technology into aging parlor infrastructure.

Short communication: Evaluation of an eating time sensor for use in pasture-based dairy systems.

The assessment of grazing behavior is important for research and practice in pasture-grazed dairy farm systems. However, few devices are available that enable assessment of cow grazing behavior at an individual animal level. This study investigated whether commercially available Smarttag "eating time" sensors (Nedap Livestock Management, Groenlo, the Netherlands) were suitable for recording the grazing time of cows. Smarttag sensors were mounted on the neck collars of multiparous Holstein-Friesian cows in a herd in Taranaki, New Zealand. Cows were randomly selected each observation day from the milking herd for 8 separate days across a 1-mo period. Trained observers conducted 90-min observation periods to evaluate the relationship between the sensor eating time measure and grazing time. A set of 5 defined cow behaviors (2 "head up" and 3 "head down" behaviors) were assessed. In total, observations of 37 cows were recorded in 14 sessions over 8 d in the study period, providing 55.5 total hours of observations. Observation data were aligned with sensor data according to the sensor time stamps and grouped into matching 15-min intervals. Interobserver reliability was assessed both before and after the main trial period, and the mean percentage eating time per observer had a coefficient of variation of 0.46% [mean 93.2, standard deviation (SD) 0.425] before and 0.07% (mean 96.3, SD 0.074) after. In the main trial, the relationship between observed (mean 70.8%) and sensor-derived (mean 69.3%) percentage eating time over the observation period gave a Pearson correlation coefficient of 0.971, concordance correlation coefficient 0.968, mean difference 1.50% points, and SD 5.8% points. Therefore, sensor-identified percentage "eating time" and observed percentage active grazing time were shown to be both very well correlated and concordant (in agreement, with high correlation and little bias). Therefore, the relationship between observed and sensor-derived data had a high degree of agreement for identifying cow grazing activity. In conclusion, Smarttag sensors are a valid and useful tool for estimating grazing activity at time periods of 1 h or more.

Short communication: Technologies and milking practices that reduce hours of work and increase flexibility through milking efficiency in pasture-based dairy farm systems.

To attract and retain quality employees, dairy farms must be competitive with other workplaces offering more conventional hours of work. Milking requires significant labor input and influences the start and end times of the working day, affecting flexibility to suit employee needs or availability. The use of labor-saving technology and milking management strategies could help with this challenge. Previous studies have used scenario modeling in attempt to quantify the value of in-parlor technologies, however, they have relied on assumptions about the effect of the technologies on labor in the dairy. Similarly, the effect of management strategies on work patterns, such as flexible milking intervals (changing the timing of milking), has not been evaluated. The aims of this study were to (1) quantify the milking labor requirements in a range of pasture-based dairy farm systems and (2) identify practices or technologies that facilitate efficient milking. A telephone survey of 500 dairy farmers in New Zealand was conducted during April and May 2018, with questions asked about milking practices and technology use. Predictive analysis showed that at peak lactation, milking required between 17 and 24 h/wk per worker for farms milking twice a day, representing 43 to 58% of a conventional 40-h work week, depending on parlor type (herringbone or rotary), the number of clusters, and herd size. Using milking intervals of 8 and 16 h (intervals between milkings), compared with the more usual 10 and 14 h, largely avoided starting milking before 0500 h. Eight percent of herds were milked once a day, which required between 7 and 14 h/wk per worker (18-35% of a 40-h week). ANOVA showed that for metrics that related to people (labor efficiency and work routine), using automatic teat spraying had a positive effect on efficiency. Having both automatic cluster removers and drafting were associated with longer milking times in terms of throughput and row/rotation time compared with using drafting only. The results highlight considerable opportunity to reduce the number of hours those milking (employers and employees) spend in the parlor and increase staff time flexibility through milking (e.g., intervals between milkings) and parlor management (e.g., row/rotation time) and use of specific technologies. This study provides useful data for those wishing to analyze the likely value of an in-parlor automation technology or management practice for an individual situation.

Field validation of protocols developed to evaluate in-line mastitis detection systems.

This paper reports on a field validation of previously developed protocols for evaluating the performance of in-line mastitis-detection systems. The protocols outlined 2 requirements of these systems: (1) to detect cows with clinical mastitis (CM) promptly and accurately to enable timely and appropriate treatment and (2) to identify cows with high somatic cell count (SCC) to manage bulk milk SCC levels. Gold standard measures, evaluation tests, performance measures, and performance targets were proposed. The current study validated the protocols on commercial dairy farms with automated in-line mastitis-detection systems using both electrical conductivity (EC) and SCC sensor systems that both monitor at whole-udder level. The protocol for requirement 1 was applied on 3 commercial farms. For requirement 2, the protocol was applied on 6 farms; 3 of them had low bulk milk SCC (128×10(3) cells/mL) and were the same farms as used for field evaluation of requirement 1. Three farms with high bulk milk SCC (270×10(3) cells/mL) were additionally enrolled. The field evaluation methodology and results were presented at a workshop including representation from 7 international suppliers of in-line mastitis-detection systems. Feedback was sought on the acceptance of standardized performance evaluation protocols and recommended refinements to the protocols. Although the methodology for requirement 1 was relatively labor intensive and required organizational skills over an extended period, no major issues were encountered during the field validation of both protocols. The validation, thus, proved the protocols to be practical. Also, no changes to the data collection process were recommended by the technology supplier representatives. However, 4 recommendations were made to refine the protocols: inclusion of an additional analysis that ignores small (low-density) clot observations in the definition of CM, extension of the time window from 4 to 5 milkings for timely alerts for CM, setting a maximum number of 10 milkings for the time window to detect a CM episode, and presentation of sensitivity for a larger range of false alerts per 1,000 milkings replacing minimum performance targets. The recommended refinements are discussed with suggested changes to the original protocols. The information presented is intended to inform further debate toward achieving international agreement on standard protocols to evaluate performance of in-line mastitis-detection systems.

Devices used by automated milking systems are similarly accurate in estimating milk yield and in collecting a representative milk sample compared with devices used by farms with conventional milk recording.

Information on accuracy of milk-sampling devices used on farms with automated milking systems (AMS) is essential for development of milk recording protocols. The hypotheses of this study were (1) devices used by AMS units are similarly accurate in estimating milk yield and in collecting representative milk samples compared with devices used by certified milk recording providers on farms with conventional milking systems (CMS) and (2) devices used on both AMS and CMS comply with accuracy criteria described by the New Zealand Standard and by the International Committee of Animal Recording (ICAR). Milk recording data from 5 AMS farms were collected during 13 milk recording test days between December 2011 and February 2013. Milk yield was estimated by ICAR-approved milk meters on AMS units. Milk samples were collected over a 48-h period and submitted to an off-site certified laboratory for milk composition analysis. Data were also collected manually from 5 to 10 cows per AMS unit; a complete milking of a cow was weighed to serve as gold standard for milk yield, and 3 milk samples per cow milking were collected and analyzed in the laboratory to serve as gold standards for milk composition. A similar procedure was used during 6 milk recording occasions with devices used during conventional milk recording at a CMS research farm. Farm type, breed, season, and region did not appear to affect accuracy of devices used on AMS units. Milk meters used by AMS units complied with ICAR limits in 12.5 and 25% of the milk recording test days for test bucket weights between 2 and 10kg and for test bucket weights >10kg, respectively. These percentages were 52 and 42%, respectively, for devices used on CMS. Analyzing all samples as one milk recording test day, 1.4% fell outside the 20% difference band for AMS compared with 1.1% of the milk samples for CMS. Devices used by AMS complied with ICAR in 73% of the milk recording test days for fat percentage, compared with 42% of the milk recording test days by devices used at the CMS farm. When analyzing all milk samples as one milk recording test day, 3.5% of the milk samples fell outside the 99% ICAR limit for AMS compared with 17.2% of the milk samples for CMS. Applying the ICAR standards for fat percentage to crude protein percentage and SCC, devices used on AMS were accurate in estimating crude protein percentage but not in estimating SCC. Thus, devices on AMS units did not comply with national nor ICAR standards with regard to milk yield and fat percentage. However, devices used on AMS were similarly or more accurate compared with devices used during conventional milk recording. It is proposed that devices used on AMS units, when calibrated regularly and when set up according to the manufacturer's instruction, have similar or improved accuracy compared with CMS devices. Because the New Zealand industry accepts data from devices currently used by certified providers for milk recording on CMS farms, results imply that the AMS devices should also be permitted to be used for milk recording.

Using activity-based monitoring systems to detect dairy cows in oestrus: a field evaluation.

AIM: To assess the use and performance of activity-based oestrus detection systems (ODS) on two commercial dairy farms using a gold standard based on profiles of concentrations of progesterone in milk, artificial insemination (AI) records and pregnancy diagnosis results. METHODS: Two activity-based ODS were evaluated in mature cows on two large pasture-grazed dairy farms (>500 cows) over the first 3 weeks of AI. Farm 1 (n=286 cows) used a leg-mounted device and cows were drafted automatically based on activity alerts. Decisions regarding AI were then made based on tail-paint and cow history for these cows. Farm 2 (n=345 cows) used a collar-mounted device and activity alerts were used in conjunction with other information, before the farmer manually selected cows for AI. The gold standard to define the timing of oestrus was based on profiles of concentrations of progesterone in milk measured twice-weekly, used in conjunction with AI records and pregnancy diagnosis results. Sensitivity and positive predictive value (PPV) were calculated for the activity-based ODS data only, and then for AI decisions, against the gold standard. RESULTS: Farm 1 had 195 confirmed oestrus events and 209 activity alerts were generated. The sensitivity of the activity-based ODS was 89.2% with a PPV of 83.3%. Using tail-paint and cow history to confirm activity-based alerts 175 cows were inseminated, resulting in a sensitivity of 89.2% and an improved PPV of 99.4%. Farm 2 had 343 confirmed oestrus events, and 726 alerts were generated by the activity-based ODS, giving a sensitivity of 69.7% with a PPV of 32.9%. A total of 386 cows had AI records, giving a sensitivity of 81.3% and PPV of 72.3%. CONCLUSIONS: The two activity-based ODS were used differently on-farm; one automatically selecting cows and the other supporting the manual selection of cows in oestrus. Only one achieved a performance level suggested to be acceptable as a stand-alone ODS. Use of additional tools, such as observation of tail paint to confirm activity-based oestrus alerts before AI, substantially improved the PPV. CLINICAL RELEVANCE: A well performing activity-based ODS can be a valuable tool in identifying cows in oestrus prior to visual confirmation of oestrus status. However the performance of these ODS technologies varies considerably.

Development of protocols to evaluate in-line mastitis-detection systems.

This paper proposes and discusses a methodology to evaluate the performance of automated mastitis-detection systems with respect to their practical value on farm. The protocols are based on 3 on-farm requirements: (1) to detect cows with clinical mastitis promptly and accurately to enable timely and appropriate treatment, (2) to identify cows with high somatic cell count to manage bulk milk SCC levels, and (3) to report the mastitis infection status of cows at the end of lactation to support decisions on individual cow dry-cow therapy. Separate protocols for each requirement are proposed and discussed, including gold standards, evaluation tests, performance indicators, and performance targets. Aspects that require further research or clarification are identified. Actual field data are used as examples. Further debate is invited, the aim being to achieve international agreement on how to evaluate and report performance of different mastitis-detection technologies. Better performance information will allow farmers to compare different mastitis-detection systems sensibly and fairly before investing. Also, the use of evaluation protocols should help technology providers to refine current, or develop new, automated mastitis-detection systems. Such developments are likely to accelerate adoption of these systems, potentially leading to improved animal health, milk quality, and labor productivity.