FW-HTF-RL: Testing a Responsible Innovation Approach for Integrating Precision Agriculture (PA) Technologies with Future Farm Workers and W ork

人类-技术前沿工作的未来(FW-HTF)项目将通过开发和测试社会和伦理上令人满意的精准农业技术和劳动力增加方法，促进对在农业中建立更强大的人机网络的基本理解。精准农业采用基于数据的农业技术和实践以及本地化的农场数据，以生成特定地点的农场建议，从而提高农场生产率和环境可持续性。为了释放精准农业的这种潜力，教育工作者和科学家迫切希望培训未来的农业劳动力。要接受任何培训，农场工人需要相信他们可以信任从这些技术获得的信息，并且需要有一条明确和可理解的途径将数据转换为有用的信息。该项目将使用南达科他州和佛蒙特州的真实农场作为活实验室，开发和测试新的精准农具(智能决策支持系统)、传感器驱动的基于绩效的激励措施以实施可持续农业实践，以及可以增强农场工人对精准农具的信任和信心的劳动力培训计划。活的实验室方法将使农场工人成为精准农具的使用者、共同生产者和共同评估者。这种互动的技术开发过程有可能增加农场工人对精准农具的信任，加强培训过程，增加农民对这些工具的采用，提高农场生产力，以及农场内外环境的可持续性。该项目的积极溢出效应还将加速共同设计和共同评估的农业人工智能创新向许多其他经济部门的过渡。该项目将使用活的实验室方法：(1)开发、部署、创建算法，并测试通过基于人工智能的综合决策支持系统将从高光谱和多光谱传感器、田间监测仪和现场养分传感器收集的数据转换为农场工人有用信息的能力；(2)试点农场内、传感器驱动的生态系统服务绩效支付机制；以及，(3)实施负责任的创新原则，以得出与政策相关的见解，以加强农业中的人机网络。这个跨学科项目团队采取的活实验室方法将：(1)促进对农业中值得信赖的人工智能的负责任创新的基础性理解；例如，在什么创新、政策和劳动力培训条件下，农场工人开始信任智能决策支持系统提出的建议；(2)开发和测试创新的智能决策支持系统，以集成来自不同来源和规模的大数据，例如无人机和现场传感器；(3)测试新型低成本纳米传感器的开发和集成，用于在活的实验室农场测量土壤和水中的磷和氮，以及(4)帮助发展对生态负责的农业的新领域；例如，基于传感器的生态系统服务付费机制如何能够彻底改变可持续人类-环境-技术伙伴关系的设计。通过教育和推广计划，该项目将培训15名跨学科的博士生，并让100多名本科生、48名农民和来自公共、私人和非营利组织的利益相关者沉浸在活的实验室中。这项研究建立在南达科他州立大学和佛蒙特州大学成功的精确农业研究计划的基础上，并设想开发新的方法来模拟这个复杂的社会技术系统，目的是成功和负责任地将农业工人转变为南达科他州和佛蒙特州乃至全国其他地区的农业工作中的数字转型。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This Future of Work at the Human-Technology Frontier (FW-HTF) project will advance the fundamental understanding of building stronger human-machine networks in agriculture through the development and testing of socially and ethically desirable precision agriculture technologies and workforce augmentation approaches. Precision agriculture employs data-based agricultural technologies and practices and localized farm data to generate site-specific farm recommendations that can improve farm productivity and environmental sustainability. To unlock this potential of precision agriculture, educators and scientists are eager to train the future farm workforce. To embrace any training, farm workers need to believe they can trust the information they will get from these technologies and there needs to be a clear and understandable path to converting data to usable information. This project will use real farms in South Dakota and Vermont as living laboratories for developing and testing new precision agriculture tools (intelligent decision support system), sensor driven performance-based incentives for implementation of sustainable agriculture practices, and workforce training initiatives that can enhance farm workers’ trust and confidence in precision agriculture tools. The living laboratory approach will involve farm workers as users, co-producers, and co-evaluators of precision agriculture tools. This interactive technological development process has the potential to increase farmworkers’ trust in precision agriculture tools, enhance the training processes, increase farmers’ adoption of these tools, improve farm productivity, and on and off-farm environmental sustainability. Positive spillover from this project will also accelerate the transition of co-designed and co-evaluated artificial intelligence innovations in agriculture into many other economic sectors.This project will use a living laboratory approach to: (1) Develop, deploy, create algorithms, and test the ability to convert data collected from hyperspectral and multispectral sensors, field monitors, and in-situ nutrient sensors into useable information for farm workers through an Artificial Intelligence-based integrative decision support system; (2) Pilot an on-farm, sensor-driven performance-based payment for ecosystem services mechanism; and, (3) Implement principles of responsible innovation to draw policy-relevant insights that can strengthen human-machine networks in agriculture. The living laboratory approach taken by this interdisciplinary project team will: (1) Advance foundational understanding of responsible innovation for trustworthy artificial intelligence in agriculture; e.g. under what conditions of innovation, policy, and workforce training do farm workers come to trust recommendations made by intelligent decision support systems; (2) Develop and test innovative intelligent decision support system to integrate big data from heterogeneous sources and scales, e.g. unmanned aerial vehicles and in situ sensors; (3) Test the development and integration of novel low-cost nano-scale sensors for measuring soil and water phosphorus and nitrogen in the living laboratory farms, and (4) Help evolve new areas of ecologically responsible farming; e.g. how sensor-based payment for ecosystem services mechanism can revolutionize design of sustainable human-environment-technology partnerships. Through educational and outreach programs, this project will train 15 interdisciplinary PhD students and immerse more than 100 undergraduate students, 48 farmers, and stakeholders from public, private, and non-profit organizations in the living laboratories. This research builds on the successful precision agriculture research initiatives at South Dakota State University and University of Vermont and envisions the development of new approaches in modeling this complex socio-technical system for the purpose of successfully and responsibly transitioning agricultural workers for digital transformations in farm work in South Dakota and Vermont, and eventually rest of the nation.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.