PFI:AIR-TT: Automating Animal Welfare Tasks to Improve Animal Health and Human Wellbeing

PFI：AIR-TT：自动化动物福利任务以改善动物健康和人类福祉

• 批准号: 1701026
• 负责人: Donald Brightsmith
• 金额: $ 19.71万
• 依托单位: Texas A&M AgriLife Research
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-06-15 至 2019-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1701026&HistoricalAwards=false
• 关键词: PFI; AIR; TT; Automating; Animal

This PFI: AIR Technology Translation project focuses on using research on the benefits of positive reinforcement training to create a digital system that improves animal wellbeing while automating tasks related to common types of animal care. This digital enrichment system is important because animal husbandry, especially tasks related to animal welfare, take a great deal of caretaker time, money, and effort that can cause caretaker burnout. Contented animals are important to multiple industries: research animals undergo significant physiological and behavioral shifts under stress, which biases scientific studies; contented farm animals have increased immune function and need less medication; conservation breeding of endangered species depends on healthy breeding animals; and ease in pet keeping encourages the health benefits of pet ownership. The project will result in a manufacture ready prototype for use with parrots that has the following features: it will offer toys and food items on a preprogramed schedule, track and log bird behavior, and run interactive "apps" that provide rewards for cognitive thinking challenges, exercise, and natural behavior expression. Parrots are an unusual group of animals, as they exist simultaneously in endangered conservation breeding, pet ownership, commercial pet breeding, and research laboratories. This proposed digital enrichment system's ability to provide many different types of stimuli at a cost of ~$300 per unit provides many advantages when compared to the $3,000-8,000 systems that exist the laboratory market and perform only limited husbandry tasks. There are basically no similar items in this primary market space, the companion bird market. Current state of the art in the market space for other species are items like monkey restraint chairs with levels and light panels, enriched poultry confinement cages (dust bath, perches, nest), and low functionality single species dog or cat products such as the iCPooch feeder.This project addresses the following technology gaps as it translates from research discovery to commercial application. The complex actions and decisions made by positive reinforcement trainers must become a set of programmed instructions carried out by the app. The system requires specialized low processing power computer vision, and computer hearing that work in a variety of environments and for a variety of species. The technology must seamlessly integrate into the home of the animal, be low maintenance, replace person-hours of work, and be based on the best behavioral science. These needs will be addressed by multidisciplinary teams of animal trainers, ornithologists, computer vision experts, computer programmers, and engineers working in the lab and aviary to create software and hardware combinations that overcome these technology gaps. The animal health and outcomes monitored will include immune function, body condition, bone density, and expression of natural behaviors. During development, the project will explore crossover applications for zoo and agricultural applications as appropriate. In addition, the project leaders will work with undergraduate and graduate students to provide them with hands-on experience working with industry partners and developing products for commercial applications. The project engages three important partners in this technology translation effort: The Schubot Exotic Bird Health Center (a research center with 150 publications on bird health), Hill Country Aviaries (housing and breeding about 2,000 birds of 60 species), and Chris Biro (a bird show owner who provides online training courses and engages several hundred thousand people in outreach annually). This team will provide the researchers with the environment for joint development, pilot testing, customer outreach, and guided commercialization to help transform this research discovery into commercial reality.

该PFI：AIR技术翻译项目的重点是利用对正强化训练益处的研究来创建一个数字系统，该系统可以改善动物健康，同时自动执行与常见类型的动物护理相关的任务。这个数字丰富系统很重要，因为畜牧业，特别是与动物福利相关的任务，需要大量的看护时间，金钱和精力，这可能会导致看护人倦怠。满足的动物对多个行业都很重要：研究动物在压力下会发生显著的生理和行为变化，这会使科学研究产生偏见;满足的农场动物免疫功能增强，需要较少的药物;濒危物种的保护繁殖依赖于健康的繁殖动物;轻松饲养宠物会促进养宠物的健康益处。该项目将产生一个用于鹦鹉的制造就绪原型，具有以下功能：它将按照预先编程的时间表提供玩具和食物，跟踪和记录鸟类行为，并运行交互式“应用程序”，为认知思维挑战，锻炼和自然行为表达提供奖励。鹦鹉是一种不寻常的动物，因为它们同时存在于濒危保护繁殖，宠物所有权，商业宠物繁殖和研究实验室中。所提出的数字富集系统以每单位约300美元的成本提供许多不同类型的刺激的能力与存在于实验室市场并且仅执行有限的饲养任务的3，000 - 8，000美元的系统相比提供了许多优点。在这个初级市场空间，伴侣鸟市场，基本上没有类似的项目。目前市场上其他物种的最新技术是带有水平和光板的猴子约束椅，丰富的家禽禁闭笼（尘浴，栖息，巢），以及低功能的单一物种狗或猫产品，如iCPooch喂食器。该项目解决了以下技术差距，因为它从研究发现转化为商业应用。正强化训练师做出的复杂动作和决定必须成为应用程序执行的一组编程指令。该系统需要专门的低处理能力计算机视觉和计算机听觉，可在各种环境和各种物种中工作。该技术必须无缝集成到动物的家，低维护，取代人工小时的工作，并以最好的行为科学为基础。 这些需求将由动物训练师，鸟类学家，计算机视觉专家，计算机程序员和在实验室和鸟舍工作的工程师组成的多学科团队来解决，以创建克服这些技术差距的软件和硬件组合。监测的动物健康和结果将包括免疫功能、身体状况、骨密度和自然行为的表达。在开发过程中，该项目将酌情探索动物园和农业应用的交叉应用。此外，项目负责人将与本科生和研究生合作，为他们提供与行业合作伙伴合作和开发商业应用产品的实践经验。 该项目涉及三个重要的合作伙伴在这一技术翻译工作：舒伯特珍奇鸟类健康中心（一个研究中心，有150个出版物的鸟类健康），山乡鸟舍（住房和繁殖约2,000鸟类的60个物种），和克里斯比罗（一个鸟展业主谁提供在线培训课程，并从事推广每年几十万人）。该团队将为研究人员提供联合开发，试点测试，客户推广和指导商业化的环境，以帮助将这项研究发现转化为商业现实。