Share on FacebookTweet about this on TwitterShare on LinkedInEmail this to someonePrint this page

The desire to automate segments of the tree fruit industry isn’t new.

At its formation, as far back as the 1960s, the Washington Tree Fruit Research Commission made examining the potential for substituting technology for labor a research priority.

by Mike Willett

Interest and effort in orchard automation has ebbed and surged over the years, largely driven by changes in industry anxiety levels regarding U.S. immigration policy and demographic realities outside our borders.

Early attempts to mechanize tasks such as apple harvest were limited by an inability to quickly pick and deliver fruit free of damage.

Also, the late 1960s marked the beginning of a mass migration from Mexico, with arrests of persons crossing the U.S./Mexico border peaking at about 1.6 million people some years between 1980 and 2000.

Those days are over. Research and development on horticultural mechanization/automation is on the upswing after decades of inactivity — and the new effort isn’t just driven by changes in the international flow of jobseekers.

Faster and more robust computing capability, highlighted by field-proven machine vision systems, has opened the door to automation opportunity.

Take, for example, Abundant Robotics. A spinoff of SRI International and now well known in the Northwest apple industry, Abundant recently announced $10 million in venture capital funding to further the development of its automated harvester.

Emerging companies, such as FFRobotics, and land grant universities, such as Washington State University and University of California, Davis, and scientists from the U.S. Department of Agriculture have also embraced harvest automation research.

These institutions and other emerging private sector firms are also seeking to expand mechanization/automation to tasks outside of harvest, including thinning, pruning and tree training, as well as the use of machine vision to estimate crop load and density.

Additionally, there has been significant interest in employing aspects of these base technologies to conduct routine orchard operations such as spraying or mowing.

Given this snapshot of automation possibilities, what should be the goal for automation/mechanization research and extension in the Pacific Northwest tree fruit industry? It’s a question the Washington Tree Fruit Research Commission looks to the industry for guidance to answer.

Many regional orchard operations have made sizable financial and infrastructural commitments to utilize the H-2A visa program for temporary agricultural workers.

How much should this investment influence thinking about where to spend grower research dollars and leverage public funding sources with a goal of automating orchard hand labor tasks?

Orchard architecture will play a large role in the success of automated harvest, given the types of end effectors currently being used to pluck fruit from the tree.

To successfully pick a high percentage of the fruit undamaged currently requires reasonably unobstructed access to the fruit. Not all current orchard systems can be easily manipulated to create the roughly “two-dimensional” canopies that appear to be necessary.

Can workable automation/mechanization innovations be applied to these systems and older orchards whose architectures pose a challenge?

How important are tasks such as fruit/flower thinning, pruning and training, which not only require the application of machine vision and substantial computing power, but also must incorporate an understanding of horticultural heuristics or rules of thumb to guide this work? What are the needs for automation of routine tasks, such as mowing and thinning?

The fruit packing equipment industry has embraced automation for handling, sorting and packing purposes, and private innovation currently appears capable of meeting the needs of the fruit handling industry.

However, is there a role for grower and broader public resources to drive innovation in this area, and what might that role be?

Equally important is an understanding of the human factors that affect and are affected by the integration of machines into work that has traditionally been performed by humans. A transition to mechanization must take into account the impact of this human-machine integration on the individual worker.

Because food production is perhaps the key societal activity, questions will inevitably be asked of fruit growers as to how automation will affect families who rely on farm work.

Almost 20 years ago, WSU and the WTFRC began the development of the first “Technology Roadmap for Tree Fruit Production.”

As we reflect on what we have learned over the past two decades and the future promise of computing intensive automation/mechanization, we want to know if we are focused on the issues of highest priority and what more we need to do. I look forward to your ideas. •

Technology research

The Washington Tree Fruit Research Commission’s deadline for new technology research proposals has been moved up to Aug. 31.

This earlier timing will allow commission staff greater flexibility in assigning projects to the appropriate research review, if a given proposal appears to be more crop-specific than broadly technology-oriented.

The request for proposals were sent July 31. The Technology Research Review will be held in early February.

– by Mike Willett, manager of the Washington Tree Fruit Research Commission, which supports applied science to improve the financial and ecological sustainability of Washington’s tree fruit industry through innovative horticulture, pest and disease management, postharvest programs, and development of new technologies. He can be reached at willett@treefruitresearch.com or 509-699-8067.