I-Corps: Predictive algorithms to determine individual feed intake in beef cattle.

I-Corps项目更广泛的影响/商业潜力是开发一种农场工具，使畜牧业生产者能够确定他们的动物吃了多少。在畜牧业经营的可变生产成本中，饲料成本约占70%。然而，很少有方法可以让生产者测量动物的采食量，而且拥有这种能力的地点数量有限，成本高昂。这种低成本的农场工具将使农民能够识别出更高效的潜在替代动物，改善他们在饲养场管理动物的方式，并量化放牧动物的摄入量。目前还没有办法在动物大规模放牧时确定牧场动物的摄入量。如果5%的美国牛肉生产商使用这个工具，那将是40,000个操作，并可能改善与50万到100万头牛相关的管理决策。I-Corps项目的基础是开发一种预测算法，利用动物的每日体重和饮水量以及天气数据来预测每日采食量。所提出的工具已经使用来自专业饲养仓的数据进行了训练，该饲养仓具有测量采食量以及动物体重和饮水量的设备。目前，最先进的系统明显高估或低估了实际采食量。所提出的工具旨在在没有昂贵的采食量系统或在不可能称重饲料的广泛放牧牧场的情况下工作。该系统已在约2200只畜棚饲养的动物和近100只小块地放牧的动物身上进行了验证，在小块地放牧可以确定采食量的地面真相。结果表明，对个体日采食量的预测准确度在92-95%之间。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact/commercial potential of this I-Corps project is the development of an on-farm tool to allow animal agriculture producers to determine how much their animals are eating. Approximately 70% of an animal agriculture operation’s variable cost of production is the cost of feed. However, there are few approaches that allow producers to measure their animals’ feed intake, and the limited number of locations that have that capacity are expensive. The proposed low cost, on-farm tool would allow farmers to identify potential replacement animals that are more efficient, improve how they manage animals in the feedlot and to quantify intakes of animals grazing pasture. Currently there is no way for pasture animal intake to be determined when animals are grazing at scale. If 5% of US beef producers made use of this tool that would be 40,000 operations and likely improve the management decisions related to upwards of 500,000 to a million cattle.This I-Corps project is based on the development of a predictive algorithm to make use of daily animal weight and water intake, along with weather data, to predict daily feed intake. The proposed tool has been trained using data from a specialized feeding barn that has equipment to measure feed intake as well as animal weight and water intake. Currently, state-of-the-art systems significantly over- or under-estimate the actual feed intake. The proposed tool intends to work in situations where either there is not an expensive feed intake system or in extensive grazing pasture situations where weighing feed is not possible. The system has been validated on ~2200 animals fed in the barn and almost 100 animals grazing small plots where a ground truth can be determined for grazing feed intake. Results have shown predictions of individual daily feed intake to within 92-95% accuracy.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.