The use of GPS tracking and the LoRaWAN network to improve productivity of grazing dairy cows

“在过去的20年里，英国每头奶牛的产奶量增加了50%（AHDB，2018）。虽然放牧草是一种廉价，可持续，高质量的饲料，但目前的放牧系统不支持这些增加的产量，因此奶牛在夏季喂养补充剂，甚至圈养它们，以确保更高的饲料摄入量。我们的愿景是开发一个系统，实时收集奶牛行为数据，并使用人工智能技术来确定何时最好地为放牧的牛群分配额外的牧场，从而提高采食量和产奶量。到目前为止，这种应用草地研究一直存在市场失灵，因为该行业很少有成熟的公司能够从推广此类工作中获得经济利益。这个创新项目将使用奶牛行为数据来跟踪放牧强度，并确定何时最好自动分配额外的放牧。这将增加产奶量，并使更多的奶牛在更多的时间里吃草，从而满足社会对奶牛吃草的需求。安装在项圈上的小型传感器使用GPS信号跟踪奶牛，并通过加速度计数据监测它们的放牧行为。牛群中的一小部分奶牛（5-8%）将在整个放牧季节佩戴项圈，收集的数据将通过低功耗广域网（LoRaWAN）发送到互联网。LoRaWAN系统（2015年）具有10公里的范围，因此它可以接收来自农场所有田地的数据，并且具有低功耗要求，因此小型电池可以在一整年内为传感器供电。它已被用于城市部门，但不是在低地农业。我们将跟踪奶牛，以确定它们何时进出挤奶厅，并监测它们在田间的放牧活动。群体放牧行为将被用来确定何时最好地分配额外的放牧。系统会产生一个信号，触发牧场大门打开，允许牛群进入新鲜的牧场。随着时间的推移，系统会建立一个关于奶牛放牧地点的数据库，用于量化每块牧场的生产力。这些信息将改善和支持牧场管理决策，如需要重新播种放牧牧场。"

"British milk yields per cow have increased by 50% in the past 20 years (AHDB, 2018). Whilst grazed grass is a cheap, sustainable, high quality feed current grazing systems do not support these increased yields so cows are fed supplements or even housed them over the summer to ensure higher feed intakes. Such actions increase feed costs and environmental problems such as slurry disposal.Our vision is to develop a system that will collect cow behavioural data in real-time and use artificial intelligence techniques to determine when best to allocate additional pasture to grazed herds resulting in higher intakes and more milk production from grass. To date there has been a market failure in such applied grassland research as there are few established companies in the sector who would financially benefit from promoting such work.This innovate project will use cow behavioural data to track grazing intensity and determine when best to automatically allocate additional grazing. This will increase milk production from grass and lead to more cows to be grazed more of the time helping to meet society's desire that dairy cows should be grazed at grass.Small collar-mounted sensors with track the cows using GPS signals and monitor their grazing behaviour through accelerometer data. A small proportion of cows in the herd (5-8%) will wear the collars throughout the grazing season and the collected data will be sent via a low power wide area network (LoRaWAN) to the internet. The LoRaWAN system (2015) has a 10 km range so it can receive data from all fields on a farm and has a low power requirement so that small batteries can be used to power the sensors over a full year. It has been used in the urban sector but not in lowland agriculture.We will track cows to determine when they move to and from the milking parlour and to monitor their grazing activity in the fields. Group grazing behaviour will be used to determine when best to allocate extra grazing. A signal will be generated to trigger a field gate to open, allowing the herd access to fresh pasture.Over time the system will build up a database of where the cows have grazed which can be used to quantify the productivity of each field. This information will improve and support pasture management decision making such as the need to re-seed grazed pastures."