Access flooding information on Disaster Resources.

November 11, 2021

TECH FOREWORD: 6 Major Learnings from Global Harvest Automation Initiative

By Walt Duflock, Vice President of Innovation

The Western Growers Innovation team is making good progress on the Global Harvest Automation Initiative as we continue to target sustainable food security by getting all the crop acres that are planted harvested.

Our target is to automate 50 percent of fresh specialty crop harvest in 10 years. For the innovators, we are developing a Common Modular Architecture as part of the Technology Stack and working with an innovator cohort of both startups and original equipment manufacturers/custom manufacturers to help startups get through development and field trials faster. For growers, our partners at Roland Berger are busy working on the first edition of the Harvest Report, which will provide a summary of the actual impact harvest automation is having and which startups are gaining traction in the market. The Harvest Report is on track to release in Q4 and will become an annual deliverable.

There are some major learnings from the harvest initiative thus far:

1.  Startups need to get the robot right. Without a robot that can do the actual work of harvest, nothing else matters. Building an 80-10-10 platform that allows startups to re-use common components for 80 percent of development, customize those components for 10 percent of development and build new innovations for 10 percent of development will help get the robot right faster.

2.  Startups need a business model with a capital strategy to support their planned pricing model and go-to market approach, as well as a machine that will provide growers with compelling economics so they will move from field trials to customers. For many startups, this issue is more important than the actual robot and both are required to scale.

3.  Harvest automation works best when it is coordinated with genetics like Bayer’s High Rise broccoli, which makes the harvest process faster and safer, and farm practices, such as the move in apples from 3D to 2D trellis growing systems which improves visibility and harvest efficiency. Genetics and farming practices can both make harvesting easier for people and robots.

4.  Startups like Burro (aka Augean Robotics) are proving that harvest aids may be as or even more valuable than actual harvest robots because Burros are here now with a plan that is working for many table grape operators and will likely work for additional crop types in the future.

5.  Venture capital (VC) is not going to get agtech where we need it for harvest automation. With long development cycles, large capital requirements and a most likely outcome of one startup/crop type, the space is not VC-friendly. WG is launching a private-public National Research Initiative with ag-friendly banks, strategic corporate venture investors, states (including California and Washington) and the federal government.

6.  Creating an online presence where the harvest community can collaborate will be a big help to accelerate harvest automation. WG is launching the HarvestWiki to help crowd source all the key grower details by crop type that startups need to know, including how large the market opportunity for the crop is; things startups need to know about growing systems and farming operations to build a successful harvester; and edge cases the startups need to know (like how to handle doubles for apple harvest). The HarvestWiki will launch in Q4 and the community will help build the content for startups, just like well-known wikis like Wikipedia.

I want to close with one request. The Roland Berger team is going to put a survey in the field to WG members. This survey is critically important to help us get data to support the first Harvest Report. It will help provide key data to support the analysis the Roland Berger team is doing to complete the report. When you see the survey request, which will be emailed to you in early November, please complete it or forward it to whoever on your team can help complete the survey with the right data. On behalf of the Roland Berger team and WG Innovation, we really appreciate the help with the survey.