Cow Health Monitor
Lead Research Organisation:
SRUC
Department Name: Research
Abstract
This project will develop a state-of-the art early-detection system for metabolic and infectious disease in dairy cattle, addressing some of the key challenges facing the UK dairy sector specifically lameness, ketosis, and acidosis. Currently yield potential (i.e. growth and milk) is hindered by the high incidence and prevalence of metabolic and infectious disease. The proposed solution will utilise animal-mounted and remote sensors to identify appropriate biological indicators or "early warning" bio-markers of the subsequent clinical disease condition. Accelerometers (neck collars) will be used to determine behavioural signatures associated with feeding behaviour (eating and rumination) and lameness (gait detection). Thermal imaging will be used to detect elevated temperature associated with local inflammation and early pyrexia (fever) indicative of infectious disease. Breath and in-line milk sensors will be used to measure, in real-time, ketones. The concentration of milk ketones correlates with a wide range of other health conditions. A ketotic state indicated by elevated plasma and milk ketones can impair immune function and thereby affect disease susceptibility
Technical Summary
This project will develop a state-of-the art early-detection system for metabolic and infectious disease in dairy cattle, addressing some of the key challenges facing the UK dairy sector specifically lameness, ketosis, and acidotic disorders/diseases. Currently yield potential (i.e. growth and milk) is hindered by the high incidence and prevalence of metabolic and infectious disease. The proposed solution will utilise animal and remote sensors to identify appropriate biological indicators or "early warning" bio-markers of the subsequent clinical disease condition. Animal sensors will determine behavioural signatures associated with feeding behaviour and lameness. Targeted thermal imaging systerms will detect elevated temperature associated with local inflammation and early pyrexia (fever) indicative of infectious disease. Breath and in-line milk sensors will be used to measure biomarker chemical constituents. Assimilation of these multiple measures will aid in early warning or diagnosis of the target conditions, and allow timely management intervention.
Planned Impact
A fully integrated system will offer services (from legacy installed oestrus detection to pre-clinical disease detection) to dairy farmers on-line over a single interface throughout the world. A number of routes to commercialisation can be established arising out of the information enabled by the platform, all with the focus of reducing the operational burden on individual farms whilst increasing productivity and profit throughout the supply chain; a feeding behaviour indicator service; information on feeding and rumination behaviour of individual animals; the range of conditions that affect these normal patterns of behaviour will be derived from individual animal-mounted technology; aligned and enhanced pre-clinical meatbolic disease and lameness detection, through novel embedded sensors deployed in a milking parlour. Thus the technology providers will enhance revenues through new product functionalities which then enable a number of services to be fashioned. The consortium comprises industries across the supply chain and each can derive commercial benefit from the outputs of the programme. Ultimately supermarket customers benefit from quality product without compromising the sustainability of the UK farming sector. All partners will continue to offer individual components as is currently the case but the real value will be derived from delivery of a fully integrated solution. The programme will de-risk the implementation but exploitation will require appropriate commercial arrangements to be established. Each partner enjoys a unique position in the supply chain and thus the generated IP can be commercially exploited in specific segments of the market. The ability to share data between different vendors on agreed royalty arrangements enables tailored services delivery. More generic outputs from the project include the capacity for large scale phenotyping of animal populations, both in terms of disease susceptibility and performance at the production level. With end-user agreements, this represents a valuable, cost-effective resource for recording traits with commercial potential to provide genetic improvement.
Livestock production in the UK will gain against its international competitors and, by retaining the IP in the UK, the economic competitiveness of the United Kingdom will be enhanced. Societal benefit will stem from improved economics and sustainability of food production and a healthier national herd. The Innovate UK programme gives the partners the platform to enhance efficiency in production and improve animal health, and this will lead top buildign of market share of new, cost-effective products. New revenue will be generated through additional services which will improve productivity, welfare, and consumer confidence. This will enable the UK companies to significantly differentiate their current product offering and hence further grow market share internationally.
Livestock production in the UK will gain against its international competitors and, by retaining the IP in the UK, the economic competitiveness of the United Kingdom will be enhanced. Societal benefit will stem from improved economics and sustainability of food production and a healthier national herd. The Innovate UK programme gives the partners the platform to enhance efficiency in production and improve animal health, and this will lead top buildign of market share of new, cost-effective products. New revenue will be generated through additional services which will improve productivity, welfare, and consumer confidence. This will enable the UK companies to significantly differentiate their current product offering and hence further grow market share internationally.
Description | The CowHealth project investigated the integration of real time sensors to provide pre-clinical detection of illness onset in cows. SRUC's work was mostly associated with a robot-milked unit established with approx. 50 cows at its Acrehead dairy. A range of sensor measurements were made and the key findings were identification of which sensors worked and which combinations of sensors provided enhanced information. Work at SRUC was followed-up in larger-scale studies based on a commercial robotically-miked unit. Initial analysis showed that collar-based eating times overestimated eating times recorded through the Hokofarm system (gold standard recording of feeder visits). Video analysis was used to augment this data and this produced a more reliable degree of correlation. Significant movements of the cow head translate to noise within the accelerometer measurements. A processing methodology was devised to counter this influence and this reduced the over estimation of eating times by around 2 hours per day. The influence on rumination times was negligible. An analysis of the outputs from a gyroscope were contrasted against those from an accelerometer and shown to be comparable when processed in the manner used currently to detect eating and rumination states. Therefore, there was no need to change the collar hardware to collect individual rumination and eating times. Based on the output of this study the rumination and eating data was enabled on farm with alerts generated based on deviation from the normal level for the animal. Afimilk found that they were able to identify the onset of various illnesses due to the decrease in both rumination and eating time for an individual animal. This functionality was released commercially. More detailed evaluation of the potential of thermal imaging to identify udder disease was conducted with 31 cows having been identified with mastitis and an equal number of healthy cows (control) used for paired image acquisition and analysis. For every mastitic and control animal, 4 thermal images of the udders were acquired on the day of detection and then throughout the recovery period (mastitic cows only). Studies confirmed that thermal imaging may be employed to detect the early stages of mastitis in dairy cows at milking and suggest that incorporation of appropriate thermal cameras in to an automated robotic milking system might provide a useful early warning system in commercial practice. However, the specification of cameras suitable for automated use in such a system, the location of such cameras on the robot and their integration in to the robot control platform will require further investigation and development. The second phase of work with the SRUC robotically milked herd provided preliminary support for the use of the Ketonix sensor to analyse milk headspace acetone as an indicator for ketosis. Work also evaluated combinations of sensors for disease identification. It was shown that a combination of collar accelerometer-based information about eating patterns could allow earlier identification of mastitis cases based on use of electrical conductivity measurements from the in-line crystalab analyser. |
Exploitation Route | Industry partners Afiilk and Fullwood are continually updating the algorithms and outputs from Afimilk collars and Fullwood robot milking/crystalab system and are progressively using project information for new alerts from their systems as follows: • Collar rumination alert - based on changes in individual animal behaviour. • Collar eating alert - based on changes in individual animal behaviour. • Collar health alerts - combination of rumination and eating alerts to identify animals which will become ill. Novel solution not available from elsewhere. • Udder health alerts - combination of rumination/eating with four quarter electrical conductivity measurements. |
Sectors | Agriculture Food and Drink |
Description | Industry partners Afiilk and Fullwood are continually updating the algorithms and outputs from Afimilk collars and Fullwood robot milking/crystalab system and are progressively using project information for new alerts from their systems as follows: • Collar rumination alert - based on changes in individual animal behaviour. • Collar eating alert - based on changes in individual animal behaviour. • Collar health alerts - combination of rumination and eating alerts to identify animals which will become ill. Novel solution not available from elsewhere. • Udder health alerts - combination of rumination/eating with four quarter electrical conductivity measurements. |
First Year Of Impact | 2017 |
Sector | Agriculture, Food and Drink |
Impact Types | Societal Economic |
Description | Digital Dairy Value-Chain for South-West Scotland and Cumbria (Strength in Places Fund) |
Amount | £21,333,299 (GBP) |
Funding ID | 99890 |
Organisation | United Kingdom Research and Innovation |
Sector | Public |
Country | United Kingdom |
Start | 02/2022 |
End | 01/2027 |
Title | CowHealth production and Health dataset |
Description | This dataset combines autoatically data from the milking Robot and data recorded by farm staff on the health Status of the cows. |
Type Of Material | Database/Collection of data |
Year Produced | 2017 |
Provided To Others? | No |
Impact | Scientists will be able to interogate the daaset for diferent relevant researc questions. |
Description | Agri-EPI Centre Ltd |
Organisation | Harper Adams University |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Assimilated into a joint bid for funding into the UK Centres of Agriculural Innovation call and were successful |
Collaborator Contribution | Joint leads |
Impact | None yet |
Start Year | 2015 |
Description | Dairy Precision Farming |
Organisation | Afimilk Agricultural Cooperative Ltd. |
Country | Israel |
Sector | Private |
PI Contribution | Our Research tem provided the Research facility interms of cows, housing facilities, techicaland Research staff and the livestock Expertise. |
Collaborator Contribution | University of Strathclyde brought in their computing Expertise while Fullwood Ltd contributed anautomatic milking machine (milking Robot).. Alfmilk brought in data capturing cplatform. |
Impact | We have identified different behavioual and physiological Parameter signatures of cows that ae related to health or commercement of disease. |
Start Year | 2008 |
Description | Dairy Precision Farming |
Organisation | Fullwood Ltd |
Country | United Kingdom |
Sector | Private |
PI Contribution | Our Research tem provided the Research facility interms of cows, housing facilities, techicaland Research staff and the livestock Expertise. |
Collaborator Contribution | University of Strathclyde brought in their computing Expertise while Fullwood Ltd contributed anautomatic milking machine (milking Robot).. Alfmilk brought in data capturing cplatform. |
Impact | We have identified different behavioual and physiological Parameter signatures of cows that ae related to health or commercement of disease. |
Start Year | 2008 |
Description | Dairy Precision Farming |
Organisation | University of Strathclyde |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Our Research tem provided the Research facility interms of cows, housing facilities, techicaland Research staff and the livestock Expertise. |
Collaborator Contribution | University of Strathclyde brought in their computing Expertise while Fullwood Ltd contributed anautomatic milking machine (milking Robot).. Alfmilk brought in data capturing cplatform. |
Impact | We have identified different behavioual and physiological Parameter signatures of cows that ae related to health or commercement of disease. |
Start Year | 2008 |
Description | European Association of Animal Science |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Keynote talk at European Conference, since led to mutiple European contacts and and interest in project activity |
Year(s) Of Engagement Activity | 2015 |
URL | http://www.eaap2015.org/ |
Description | Invited talk "On the use of multiple sensor technologies in monitoring udder health status in dairy cows" |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | I was nominated as an invited speaker for the Italian Society of Buiatrics day meeting on the 12th October 2016. The main thrust of the meeting was to discuss mastitis. This provided a good opportunity for presentation of some of the work and outputs in the Cow Health Monitor project, and use of multiple sensor technologies to be able to not only identify clinical animal cases, but also the animal in the process of becoming a clinical case. |
Year(s) Of Engagement Activity | 2016 |
URL | http://www.buiatria.it/index.php/giornate-buiatriche |
Description | Open day |
Form Of Engagement Activity | Participation in an open day or visit at my research institution |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | Open day at SRUC Barony |
Year(s) Of Engagement Activity | 2018 |
Description | School visits |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | Local |
Primary Audience | Schools |
Results and Impact | Schoo visits |
Year(s) Of Engagement Activity | 2017 |
Description | Use of technolgy in cow management |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Gave a talk at BSAS |
Year(s) Of Engagement Activity | 2017 |