TechnologyBackground: The MilkHub system grew out of the dairy industry’s need for an efficient and effective mastitis detector. Rather than taking an incremental development approach based on existing agri-tech products a different systems approach was adopted drawing on skills and experience from a telecommunication product development background. The key aim at the outset was to develop a practical mastitis detector. It was felt that any unnecessary complexity, maintenance or user interaction would inevitably lead to the equipment falling into disuse. This key aim set the direction for the design of the technology and of the user interface. It was decided to implement the physical design as a flow-through system with no moving parts, no milk to sensor contact, no reagents, and no sampling or mixing. This was considered essential for a zero-maintenance, clean-in-place sensor. The research literature was extensively reviewed and industry experts consulted to identify possible mastitis indicators that were candidates for inclusion in the new detection system. The review turned up the usual suspects with many associated qualifications and limitations. These included: conductivity, yield, temperature, colour and density. The science effort turned to understanding the qualifications, overcoming the limitations, and extending the effectiveness of these or derived indicators. The technology effort focused on making the required measurements accurately within the physical constraints. Always “on” automated operation was chosen as the best mode to minimise the need for user input. The idea was to identify and rank cows by infection severity and allow for drafting during any milking via a single step user request. |
| Implementation The technology solution includes a flow-through manifold with high-flow and low-flow pathways. Precision measurements are made on the low-flow pathway where the milk is free of entrapped air. Patented high frequency electric field techniques together with other sensors are used to get instantaneous measurements through the plastic containment walls. These measurements are used separately and in combinations to obtain the characteristic signature of every cow as it milks. These signatures are stored in an embedded database for every cow every milking. The stored signatures are processed to detect levels and changes indicative of infection, specific to each cow milking by milking. Indicators of this type are combined and normalised to the herd as a whole to provide a ranking of worst to least infected cows. While the initial effort was directed towards mastitis detection it was recognised that the sensor measurements and database provided a powerful information set to determine other parameters important for dairy management. These have been developed in parallel with the mastitis detection feature, to give a combine system capability including:
|