Objective: Meet with Penn State University members of the project and visit apple growers and an apple packing facility.

 

1. Bear Mountain Orchards

Met with Joy Cline, orchard manager. BMO has 1,000 acres of fruit trees, of which 480 acres are apples and 280 acres are peaches. About 70% of the apple trees are already in the vertical axis (“tree wall”) architecture. This percentage is expected to increase to 100% in the future. On their tree walls BMO uses M9 root stock grafted with many different varieties of apples. The tree spacing is 6 x 16 ft. The trees are not expected to grow more than 4 ft. to the sides so all fruit get plenty of sunshine. On the traditional tree architecture the spacing is 6-7 x 16-19 ft. and the root stock is M26. Apple harvest takes 45 people, separated in four groups. Three groups are paid by volume  to pick apples for processing; one group is paid by hour to pick apples for the fresh market. This allows the pickers working on the fresh market apples to be careful with the fruit and avoid bruising. This is a challenge to any automation solution, which would have to meet the (low) human bruising standards. Joy suggested that we first develop a thinning-assistive device rather than a harvest-assisting device, since during thinning the fruit is simply let to fall on the ground.

 

2. Adams County Nursery

Met with Chris Baugher to see the Blueline platform. The platform has a crude mechanical-feedback device to keep it near one side of the tree row. It only works for tree walls, not for traditional tree architectures. Thinning and pruning are the two activities that have benefited the most from the platform. The tree walls are held in place by thin (1/2”?) metal rods rather than thick (1”?) wooden poles as at BMO. They have old tree walls that they want to make narrower so they can mechanize it.

 

3. Hollabaugh Bros.

Met with Bruce Hollabaugh. They are trying for the first time a high-density tree wall with spacing 4.5 x 14 ft. Problem: vehicles cannot turn around inside row, must go all the way to the end to maneuver. The Hollabaugh’s don’t use ladders and limit their tree sizes to 7.5 ft. Bruce mentioned that one of the obstacles to the introduction of new automation technology is the legacy of all the equipment that growers already have. Different growers own different sets of equipment, suited for their row architectures (especially row separation). Every new automation technology would have to be studied and introduced on a case-by-case basis. Another challenge for automation: some apple varieties are more robust to bruising than others.

 

4. PSU Fruit Research and Extension Center

Met with Henry Ngugi. Henry presented the research on pest and tree disease detection and prevention, including the experiments being done at the FREC greenhouse.

 

5. Adams County Nursery

Met with Phil Baugher and visited the apple tree nursery. This field has 10,000 trees. They are grafted in the field and sold after two years of growth. Trees are dug in September and sold in January. During these four months they are counted, measured, sorted, and stored in a warehouse. Sorting is by diameter: 3/8”, 1/2", 5/8”, 3/4". Trees are sold one year before they are actually shipped, and by the time of sale the nursery does not have an estimate of how many trees of each diameter they will have available to deliver. They cannot measure tree diameter in the field because it is prohibitively expensive to measure their 700,000 trees. (Nurseries in CA have 1.5 million trees.) If they could do it, they would be able to control inventory much more accurately. Tree separation is 12” x 42”. Main challenges to diameter measurement are double trunks and thick weeds (see pictures at the project AFS space). Asian pears, for example, are very sensitive to weed killers and therefore weed sprays have to be used sparingly. Tree diameter is measured 1.5” above graft and in the direction of the largest semi-axis. The nursery runs a John Deere 6000 (not sure what machine this is) which we could use to install sensors.

 

6. Rice Fruit Company

Met first with Daniel Rice and Lee Showalter, and later with John Rice (CEO). Rice packs 1.5 to 2 million boxes of fruit annually, mostly apples plus pears, peaches, and nectarines. They use a 2-step process: a bulk characterization of fruit by size and color, after which the fruit are stored in an atmosphere-controlled warehouse for up to one year before being sold. The sorting process is almost entirely automated, using machinery from New Zealand that handles 23 tons of fruit/hour. John Rice mentioned that fruit bruising is the single most important factor to impact fruit value. Mechanical harvesting never worked because apples get bruised if they fall on a hard surface from 3”. Harvest takes some 10 weeks of the year, and requires specialized labor. USA apples compete with fruit from Chile, China, and New Zealand. Chinese growers have no credible system for disease inspection. Mexico will start buying apples from China, which could end up in the US.

Rice would like to be able to detect internal breakdown during storage, which is invisible to the customer until he/she buys the fruit. Breakdown detection is not a trivial task since humans cannot do it, and therefore there is no process to build from. Rice imagines that vision technologies (especially IR) could address the problem.

Food traceability is starting to become an issue. RFC can trace an apple back to the orchard it came from but not to the individual blocks (not to mention individual trees) in the orchard.