The term ‘Smart farming’ is most readily associated with the advent of tools and technologies to support the agricultural sector, and these come in a wide variety of forms. They usually encompass a sense or measurement component, aligned with data capture and assimilation to an informative output. A relatively simple example might be Electronic Identification, where a transponder is embedded in a plastic eartag. This may be energised by an external reader device, to transmit a unique identifier code that relates directly to the UK ID number, say in cattle. This data may be used to inform for animal shedding or other identification purpose, sometimes in combination with another system.
Beef farming is starting to see technologies come into the sector. At the abattoir, video grading systems have gained acceptance in other territories over the past decade, and they are now being installed in the UK. These systems are certified to be used to report the obligatory classifications to the EU. They also provide additional estimates of yield, with in-built prediction models for individual primal yields. The E+V VBS2000 system has been trialled and subsequently certified in the UK to report classifications.
Accurate and non-destructive measurement or prediction of meat eating quality is still the ‘holy grail’ for the processing and retail sectors. Systems have been tested using physical/optical modalities to estimate the quality of the meat, both sensorial and nutritional. Some limited success has been had using near infra-red optical methods, but routine commercial deployment at the abattoir has not occurred to date.
Back at the farm, there are aspirations to provide estimate measures of similar characteristics on the LIVE animal. Systems are currently under development, which use camera technology, originally derived from the gaming sector, to provide 3D topography of animals, and use this to estimate a number of parameters relating to yield (Innovent Technologies Ltd). Underpinning these systems are modular autoweighing platforms that integrate weighing at the water trough (Ritchie Implements Ltd). These have been tested against calibrated weighing systems, which showed no difference in computed ADG between the two systems (Ross et al BSAS 2016). They are now commercially available, and will offer instant growth information at the individual animal level, to the farmer, via computer or a phone application.
Even simple weight information can give indications of individual animal health and performance, by computing time series trends, from daily average weight. Any individual deviation from expected trends will alert that there is a cause that is worth investing management/inspection time from the stockman perspective, by focusing on the individual animal.
Other behaviour monitoring devices are starting to be used in the beef sector. As a means of improving genetics through AI and avoid synchronisation, natural oestrus may be detected by leg and neck mounted devices, to allow optimal AI. Some of these devices have multiple uses, for example one has the ability also to detect eating and rumination bouts (Afimilk/Silent Herdsman collar). Using these outputs, early signs of animal health and efficiency issues are possible to be picked up. The outputs are conveniently displayed on computer system, so the stockman can review individual livestock over a time period. Others can also record lying/standing/walking behaviours, with potential for lameness detection (ICE Robotics). These data can provide valuable information on behaviours that relate to health issues.
Other behaviour monitoring devices may be attached elsewhere on the animal such as the tail. Monitoring tail movements over time can give a robust indication of calving time. There are commercially available variants in the market. A recent study showed that a model derived from raw data on a tail-mounted activity system, could give prediction sensitivities and positive predictive values of above 0.96 in both cases (Ross et al BSAS 2015). The key is in measurement of dystocia, and not merely the detection of the parturition point. This is where a combination so multiple data streams may benefit the prediction accuracy. Indeed work is on-going in this respect, using sensor combinations to identify the key parameters relating to dystocia.
Finally, all the systems above can add value to the production chain, whether it is a producer or processor. Indeed incorporation of these systems may reveal step changes in management practices that offer significant improvement in production efficiencies, and these will be the primary source of benefit. The secondary benefits may be in the provision of a rich source of routine phenotyping data to complement genetic/genomic improvement programmes.
Agri-EPI Centre (www.agri-epicentre.com), one of the UK’s new Centres of Agricultural Innovation is driving forward these and other initiatives by deploying these types of technology on a ‘satellite’ farm network consisting of commercial farms implementing these technologies, for both evaluation and industry demonstration purposes.