CORDIS - Forschungsergebnisse der EU
CORDIS

The Development of a Next Generation Robotic Milking Parlour

Final Report Summary - ROTABOT (The Development of a Next Generation Robotic Milking Parlour)

Executive Summary:
The Rotabot SMEs and the Exploitation Manager (Dr. Edmond Harty) are very sensitive about the results generated in the project and they do not wish their competitors knowing what they are working on. Therefore they have requested not to provide a publishable summary, other then the press release issued below and the presentation about the project given by the project Co-ordinator Dr. Joseph Walsh at the 8th International Conference on Sensing Technology (ICST 2014) in September 2014. There was also a request by a suspected competitor of the lead SME Dairymaster under EU Freedom of Information Legislation for information relating to the Rotabot project (Proposal, PR Reports and Deliverables).

“The Development of the Next Generation of Milking Parlour at IT, Tralee”
The Institute of Technology, Tralee has recently been awarded a research grant of €1.6 million to work with partners across Europe in the development of the next generation of milking parlour. This project is supported by the European Union through the 7th Framework Programme (FP7) “Research for the benefit of SME’s”, (Grant agreement number 315407) and will be coordinated by IT, Tralee. The project consortium is comprised of research partners and SMEs from five different European countries. The research partners are from Ireland (IT, Tralee), Germany (Fraunhofer-esellschaftzurFördererung der AngewandetenForschung E.V.) and Spain (Innovacio IRecerca Industrial I Sostenible SL). The three SME partners are from Ireland, United Kingdom and the Netherlands. The collaborative research project has officially started on October 1st 2012. The project will last for two years and the goal of the project is to develop a working prototype of the next generation of milking parlour for testing and demonstration purposes. The project coordinator Dr. Joseph Walsh IT, Tralee said ‘with over 80% of the labour on European dairy farms predominately provided by the farming family, and with the increasing need for on-farm productivity improvements to meet the competitiveness needs of a global agricultural produce marketplace, this project will help ensure the sustainability and profitability of the family farm model for years to come.’

Project Context and Objectives:
The main objective of the Rotabot project is to deliver an intelligent next generation robotic milking system incorporating autmated teat cleaning, cluster attachment and post milking disinfection on a scalable rotary milking installation. The system will be suitable for use in both new and existing rotary milking parlours.

The scale of dairy farming throughout the world and in Europe has changed significantly over recent years, with a move towards larger (200+ cows) enterprises which are profit driven. This is primarily due to world market pressures and the reduction in liquid milk prices. This step change has resulted in a demand for technologies and equipment which can reduce costs and labour inputs with concurrent increases in capacity to provide economies of scale. These economies of scale however will not yield profit unless quality is maintained and this is a central focus of any new technologies developed for the sector. Currently within the EU there are approximately 13.7 million farms of which 2.5 million are dairy farms. Of those 2.5 million 5% or 126,000 have a herd size greater than fifty, 93,000 between fifty and one hundred cows and 33,000 greater than one hundred cows, with over 2 million farms with less than 50 cows. Of these dairy farms, 3% - 4% invest in new milking parlours systems every year, which generates a substantial market for automated rotary parlour sales. Robotic systems are currently employed in a number of industries where repetitive tasks can be mechanically automated and controlled by a computer. The challenge to achieve this in the milking process is that the automated mechanical system has to interface with a live animal, the cow. Therefore the robot needs to adapt itself to a continuously changing set of parameters. This is very unlike most robotic processes, where the target for the robot is usually at a fixed position and the interactions are with standardised entities. Robotic milking is currently in use in Europe particularly in farms in the Netherlands, Germany and Denmark where traditionally the farm sizes are small with a maximum number of 80 cows. Most major milking equipment companies now have a robotic milking machine in their product range. These systems can be either single or multi stall, where the price per stall is approximately €120,000. Currently 60 - 70 cows can be serviced by a single stall over a 24 hour period. It takes on average up to two minutes for the robotic attachment of the cluster on each individual cow and the system must then wait (~ 5 minutes) for the milking operation to finish before the next cow can enter the stall. In conventional manual milking systems, a human can normally clean the udder and attach the cluster in less than 10 seconds. With the cost of the current robots on the market and the excessive cluster attachment time, it is clear that these Voluntary Milking Systems (VMS) robots are not suitable for the number and scale of commercial farms, where hundreds or thousands of cows have to be milked on a daily basis. Therefore current commercially

Project Results:
WP2 Project Requirements involved a detailed review of patents and of updated state-of the-art technologies to ensure the project objectives were advanced using the most up to date technology. This involved a review of current automatic milking systems, vision systems and robotics, and identified any changes that may have occurred in technologies since the project proposal. The main result of this work package was a review of the project objectives to ensure that the SME’s objectives had not changed due to the availability of new technology, also an extensive list of technical requirements were developed during this work package.

WP3 Design and development of the Imaging system this work package was an RTD activity and involved the design and development of an imaging system for cow teat recognition and tracking, that is suitable for robotised rotary milking installations. It has to be able to operate in a harsh farm environment and to keep pace with the high speed of the robot developed in WP 4 used for teat cup attachment. This work was led by IRIS working in collaboration with IPA and the other project partners, and ran from the start of the project until month 20. The work undertaken developed a list of different sensor technologies that could approach the technical task of teat recognition and returning the coordinates of said teat. Based upon set criteria of timing, tolerance and pricing, the most cost effective and reliable technology was chosen. The results from this work package underpin our choice of detection system that will not only give the teat location to the robot arm but also aid the arm in its location with respect to the cow. The development of a mock-up cow was significant in assisting the development of the algorithm for teat detection and highlighting the difficulties in doing so. It also aided in showing what the scene would look like when the respective optics were chosen. The results of image recognition of the teat from the 2D camera and the initial developmentof the 3D teat detection algorithm were significant.

WP4 Robot Design and Test, this workpackage developed a process and requirements analysis of the system. Different kinematical layouts were developed and compared to commercial available solutions. The most suitable solution was selected and elaborated in detail. The main result of this work package was the conceptual layout and the design of a custom made robotic system for grabbing teat cups and attaching them to teats. The mechanical structure of the arm and the carriage moving along with the parlour was determined. The main drives were selected and ordered. The complete system was simulated. During this period a control system for the custom robot was developed and linked with the teat detection and tracking algorithm. The final parts were manufactured and a prototype was built up. The custom robot was tested in a lab environment and a demonstration was made demonstrating the functionality of the system.

WP5 System Integration involved the control and integration of all developed systems. The communication protocol between vision and robotic system were investigated. The main result achieved is that the main controller for both custom and backup solution has been selected and ordered. A system for effective teat cleaning was also developed, tested and manufactured.

WP6 This work package described the evaluation process of the Rotabotic systems by way of including the system test reports during the development of the integration of the system. The integration of the vision system with the robot as described in WP5 was evaluated as updates were provided by the relevant RTDs. ITT performed a number of tests to evaluate the performance of the system, the results of these tests were distributed to the RTD members of the consortium and steps to improve the system were implemented. The validation began by testing the communication protocol between the vision system and the robot. It was followed by testing the quality of the detections of the vision system using an artificial udder. The tests followed the development of the integration closely from testing the communication protocol to detecting a single teat to tracking a single teat and to eventually detecting and tracking all four teats of the udder.

WP7 The demonstration of the RotaBot system took place on October 1st, 2014 on site at Dairymaster, Causeway. The demonstration took place in a laboratory environment and consisted of demonstrations of the ABB industrial robot integrated with the vision system in a number of different scenarios and also of the custom robot developed and built by Dairymaster and controlled by software developed by IPA. The custom robot was also successfully demostrated.

WP8 Knowledge Transfer and Dissemination was active for the entire duration, i.e. 24 months of the project. The work package deliverables, with the exception of the project website were dynamic and evolved with the research outputs of the project. For the current reporting period there were six tasks of relevance, Task 8.2 to 8.7 inclusive. Within PR2 there were three scheduled deliverables, D8.3 Project Results Video Clip, D8.5 Final Plan for use and Dissemination and D8.6 Plans for Product Manufacture, which have been achieved and submitted to the REA on schedule.

Potential Impact:
The main impacts of this technology are as follows:
1. Launch new opportunities for consortium SMEs.
2. Increased productivity through modernisation and automated systems.
3. Sustainability and security for the future of the modern dairy industry.
4. Increase in exporting power and economic benefits within a more productive agricultural industry.
5. Food security.
6. Improve herd health by reducing infection.
7. Improved quality of life for dairy farmers.
8. Planning Issues Relating to expansion.

1: Launch new opportunities for consortium SMEs
In general given current economic conditions, herds of 150 cows or more are more profitable enterprises, with the most efficient parlour technology. Herds under this size should, if labour is cheap and readily available, minimize their parlour investment and use low cost parlour options such as swing parlours and flat barns. As already discussed in the previous sections the trends in farm size show increasing herd numbers, therefore the potential market higher-end parlour technology will grow substantially in the coming years. These figures show the scope for consolidation into larger efficient semi-automated enterprises.

2: Increased productivity through modernisation and automated systems
The European indicators show that dairy farms are getting larger, between 2005 and 2007 the number of dairy farms with a herd size of greater than 100 cows increased by 10 % where as the number of dairy farms with a herd size less than 50 fell by almost 12.5 %. Not only did the farm size increase but the average yield per cow also increased after the introduction of the milk quota as farmers sought to fill their production quota in as an efficient manner as possible. The increases in automation and the availability of a high level of herd data allows for the greater use of precision agricultural techniques, to continue the increase in average yield per cow/yield per unit input cost.

3: Sustainability and Security for the Future of the Modern Dairy Industry
If this technology is widely adopted it will position the EU agricultural industry at an international forefront. This technology will allow the continuity of generation and the sustainability of family way of life including rural demographics. From an environmental perspective, existing robotic milking installations cannot compete with rotary installations in terms of energy usage in vacuum pumps, cooling, cleaning, water disposal etc. For example, the uses of electricity in a milk parlour are milk cooling 37% water heating 31% vacuum pump 19% and lighting 10% the remaining 3% is made up of washing, feeding and air compressors.

4: Increase in Exporting Power and Economic Benefits within a more Productive Agricultural Industry
Since the early 1960s, consumption of milk per capita in the developing countries has almost doubled. China in particular has seen per capita consumption of milk increase tenfold, whereas Brazil has seen increases of almost 40%. Dairy product exports are a key economic factor within the EU. The development of this new technology within a European context will allow Europe to become the number one market for automatic rotary parlour production and exportation.

5: Food Security
As the consumption level of dairy produce is increasing in developing countries, the implementation of dairy product food security is paralleled with the abilities to export while also able to meet local demands. Automating the industry through the proposed application not only improves productivity but also quality. This system will also allow for better long-term projections of yields and exports, because of the increased time available to the farm manager and the reduced likelihood of milk quality issues.

6: Improved Herd Health by Reducing Infection.
It has been proven that pre-milking cleaning improves quality and post milking teat disinfection prevents infection. As hygiene procedures can also be automated by such a robotic rotary system, (teat cleaning and disinfection) hygiene levels will also be increased. Lower Somatic Cell Counts (SCC) due to decreased incidences of mastitis also means longer shelf life, higher milk quality, lower processing costs and lower milk distribution costs. Within the Rotabot system the milking process is consistent for every cow and every visit, and is not influenced by differing hygiene standards of different operators milking the cows thus improving udder health. Lower milking and having a quality control system to ensure every teat is clean times allow more time for farm management which allows for the implementation of better hygiene standards, which in turn leads to better animal health and milk quality.

7: Improved quality of life for dairy farmers.
It is widely recognised that the use of greater automation in the milking process allows for larger farm sizes without increasing the work load of the farm operators. Almost half of the farms in the Netherlands and Denmark install automatic milkers when updating milking parlours. The primary reasons for the keen adaptation to technology are to improve quality of life and overcome labour shortages. This newly introduced automated process has had tremendous benefits in terms of freeing up time for farm management, supervision and allows parlour down-time during which no personnel are required on farm. With more time available for management, better health & safety standards can be set, with the added bonus of less operator fatigue leading to fewer mistakes and accidents. Shorter milking times also lead to a more socially acceptable way of life, which is vital if future generations are to be encouraged in to farming. Dairy farmers can suffer from Repetitive Strain Injury (RSI) which is an injury of the musculoskeletal and nervous systems that may be caused by repetitive tasks, forceful exertions, vibrations, mechanical compression (pressing against hard surfaces), or sustained or awkward positions. The Rotabot would remove a number of the risks associated with RSI in dairy farmers.

Large scale megafarms are not the target market for the Rotabot. The demand for a technology such as Rotabot comes from farms of 150 – 300 cows which are run as family businesses and where labour is supplied by the family. The SMEs think that mega farms will not be the future for Europe. The SMEs sees an increase in farms of 150-300 cows that want to automate because they want to keep the farm in the family for as long as possible and they want to decrease their cost price per litre milk. It is also important for these farmer families to let the robot/automation take over the physical labour. As the farmer himself is also the milker, he is the most important person for SMEs. The Rotabot will reduce the burden on the farmer and family and it is our goal to make the Rotabot cost effective so as to be economic at this herd size. It is very important to make both systems affordable for the family farms, so that not only mega farms can profit on it. Try and make this project accessible to all. The real turnover will come from the family farms with 150-300 cows.

8: Planning Issues relating to expansion
A more efficient rotary palour will solve a number of planning issues relating to farm expansion, as countries such as the Netherlands currently have restrictions relating to new farm building being confined to the footprint of pre-existing buildings. The increased efficiencies provided by the
Rotabot will allow expansion in farm capacity without the need for larger farm buildings.

List of Websites:
Project public website address: www.rotabot.eu

For further information contact:

Dr. Joseph Walsh
School of Science, Technology, Engineering and Mathematics
ITTralee – South Campus
Clash,
Tralee,
Co. Kerry

Phone: +353 66 7144215
Email: joseph.walsh@staff.ittralee.ie