Periodic Reporting for period 1 - AGRICAM (AGRICAM – reducing the usage of antibiotics and increasing animal welfare through advanced thermal imaging system for detecting early cases of mastitis in dairy cows)
Periodo di rendicontazione: 2019-08-01 al 2020-01-31
Mastitis, a painful inflammation in the dairy cow’s mammary glands, is one of the most wide-spread diseases in the dairy industry, inflicting serious damages on a global scale. In Europe alone, mastitis costs the farmers more than 60 million EUR per year1 and causes very large losses in milk production and degradation in milk quality. In herds without an effective mastitis control program, approx. 40 % of cows are infected every 6 months. Antibiotic usage is currently the only solution for treating mastitis, however, in many cases, this method is ineffective and leads to developing bacteria that are resistant to antibiotics. On top of that, global antibiotic use in food animals outweighs human consumption by nearly three times and will rise 53% globally between 2013 and 2030. This scale-up in antibiotics will be devastating to efforts to conserve its effectiveness. Also, because of the inappropriate use of these drugs in animals is a leading cause of rising antimicrobial resistance (AMR). For many years, scientists all over the world tried to find a non-antibiotic treatment against mastitis. Agricam AB has developed a revolutionary technology capable of addressing this problem - a sensor-solution to monitor lactating cows. Our patented heat camera systems (TRL-8) perform over 10 000 analyses daily for large scale and heavy-duty dairy farms in Sweden, monitoring and preventing mastitis amongst the herd 24-7. Our innovation emerged from military Infra-Red camera technology and became a tool to combat mastitis in dairy cows. To-date, Agricam has been installed at 30+ dairy farms in Sweden and one in Russia. The technology has proven its capacity to decrease the use of anti-biotics, bringing measurable economic benefits for the farmers. During this Phase 1 project, we will develop a roadmap to bring Agricam to the European markets.
During this study, we conducted four independent pilots to assess the technical and commercial feasibility of the Agricam thermal camera system. These pilots involved the Agri-Food and Biosciences Research Institute in Ireland and three small dairy farms in Sweden. The pilots led to key insights to guide the future development of our business. These collaborations highlighted important commercial barriers for the large scale adoption of the Agricam system both in terms of pricing and logistics. First of all, the current price of the system is beyond the purchasing power of small farms, which manage the majority of dairy products in the EU. Furthermore, one of the key aspects of the system, its capacity to detect mastitis pre-symptomatically, makes it difficult for farmers to see the value-added of the system. In contrast, thanks to our work with Agri-Food and Biosciences Research Institute the potential of Agricam as a research tool became apparent. Indeed, this led to an agreement for a three-year collaboration starting in March 2020 to conduct a study on herd health management. Another insight derived from this study was the urgent need for fast on-site pathogen detection. Participating farmers stressed the importance of having access to this information in a timely manner and at an affordable price. Currently, the procedure to find out which bacteria is causing the mastitis in the herd is slow and costly. Not only due to the time it takes for the bacterial growth to form, but also due to the time, it takes to ship the sample to a laboratory and waiting for it to be assessed. Furthermore, the bacteriological analysis of a milk sample costs between SEK 165-400/ EUR 15-40 (ex VAT) depending on the analysis (based on SVA's price list 2018). Such cost is very high for owners of small farms and by the time the results are back the infection is likely to have spread in the herd. Agricam’s team has the combination of skills to develop a system with the capabilities to satisfy this unmet need. We have developed Bacticam a system that enables farmers to obtain bacteriological information in 8-12 hours and half the cost of current tests. Bacticam consists of a sampling and heating cabinet where bacteria are cultured on a petri dish and then analysed using a machine learning-based image classifier. The potential of Bacticam has been recognized by our partners some of which have committed to collaborate in a future EIC Accelerator project. We have conducted a careful study of the market potential for Bacticam and the other products in Agricam’s portfolio and there is a positive forecast for the following years.
The accurate identification of mastitis pathogens is critical to guide therapy. However, current laboratory protocols are not able to provide a convenient solution for farmers. Procedures are costly and turnaround times are lengthy. The latter prevents farmers from taking action swiftly and prevent the spread of the infection throughout the herd. The development of systems for on-farm identification of mastitis-associated pathogens is essential for quick therapeutical intervention. I recent years, several multiple-media culture systems have been developed for specific on-farm diagnosis. Solutions like the Minnesota Easy Culture System Tri-Plate, Accumast and Quad-plates have shown their potential for diagnosis of specific mastitis-related microbes. These systems rely on the segmentation of cultures into different chromogenic media. Currently, culture results must be interpreted by the user using visual inspection. Experienced readers can correctly identify specific mastitis agents with 86% of the agreement to standard laboratory methods.10 Success rates for untrained professionals are obviously much lower. This is where Bacticam can revolutionise on-farm pathogen identification. By leveraging on recent advances in Machine Learning methods for image classification, we have been able to automate the pathogen identification. Bacticam’s performance is bound to grow as data fed into the algorithm increases. By the end of the Phase 2 project, we expect to surpass the performance of trained professionals – bringing expert-level performance to every farm.