Information technology is remaking every sector of the economy, and agriculture is no exception. In fact, agriculture may be the most vital segment — we all need food, and it must be produced as productively and sustainably as possible to feed an expanding world population. This is all happening in the context of growing world hunger — an estimated 690 million people go to bed on an empty stomach each night, according to the United Nations World Food Programme.
Technology is bringing tremendous advances to the practice of agriculture. For example, take the use of sensors and geographic information systems for precision, site-specific crop management, in which data are continuously collected directly in the fields using unmanned aerial systems (UAS) or satellite/aerial mounted sensors to monitor variables like soil nutrient status, moisture content, extent of pest infestation, etc.
What we need to capitalize on this and other technology infusions are innovations in education — at scale — to train students of agriscience in the new data science skills that will help revolutionize agriculture. Termed “ag informatics,” these capabilities include software coding, machine learning, data visualization, data wrangling (manipulation), and data interpretation.
For example, at an “Ideas Engine” collaborative event organized by the U.S. Department of Agriculture-National Institute of Food and Agriculture (USDA-NIFA), participants overwhelmingly voted in favor of training agriscience students to “combine agroecosystem knowledge with analytical skills in machine learning and data mining; statistical and quantitative analysis; data visualization and problem solving.”
That drives home the need to develop, implement, assess and disseminate courses in ag informatics to instruct agriscience students in computational thinking and software development. We refer to computational thinking as skills related to data analysis, model development and decision making. Helping students become well-versed in the evaluation of data quality, appropriate analysis techniques, knowledge extraction, and ultimately data-based decision making are all part of computational thinking.
These abilities are equally applicable to data generated in field/indoor crop production, food processing, livestock management, forestry, environmental sciences, economics, and health sciences. Agriscience students gathering data and performing research can use software to build complex, multi- dimensional analyses and algorithms that are not practical via traditional curriculum.
A Higher Education Challenge (HEC) grant from USDA-NIFA will help Purdue Engineering develop new courses in ag informatics to prepare agriscience graduates for practical, data-driven work in the agricultural industry, on individual farms, and in academia. We are creating a two- week course module that will be integrated into 12 existing courses offered to undergraduates at Purdue University and partners, the University of Kentucky and Tuskegee University. Three semester-long courses also will be designed, with a goal to build career-specific computational thinking and software development skill sets to increase students’ competencies for ag informatics jobs.
To scale these innovations beyond the courses, we will cultivate open-source communities in agriculture through experiential learning; professional development opportunities; and development and release of educational materials like videos, code and libraries via the Brightspace learning management system (LMS). The aim is to help students enhance networking skills, promote teamwork, and develop readiness for ag informatics careers.
Another of our inventive approaches is to encourage agriscience students to work in pairs while learning coding. The students will progress significantly faster than if they worked in isolation, and they will develop more confidence in their abilities to see a project through from conception to delivery of viable product and beyond. Paired students also will experience a true sense of community.
Although a few land-grant universities have introduced data-driven, agriculture-related certificate programs for graduate students, opportunities for undergraduates in the agriscience disciplines — including agricultural and biological engineering, agricultural systems management, and agricultural economics — are limited. We are excited that the HEC grant allows us to build up computational thinking and software development skills for both undergraduate and graduate agriscience students as they prepare to be leaders in industry, manage rapid digitization of farms and the food chain, or take up innovative projects as graduate students in academia.
Dharmendra Saraswat, PhD and Fellow, Indian Society of Agricultural Engineers (ISAE)
Associate Professor, School of Agricultural and Biological Engineering
College of Engineering, Purdue University