Skip to main content Skip to navigation

Implementing Robotic Harvest Technology

View Print Version


In the early stages of implementing a technology that the WTFRC was founded to develop in 1969, it may be time to contemplate all the current and future opportunities to further automate the apple and pear industries. It is time to start identifying synergistic technologies such as mechanical pruning, 10 foot drive rows, vertical canopies, handling properties of varieties and growth habit of the trees.

Status, March 2016: The technology is able to remove apples (and pears) at a quick pace with no damage from the process. The machine is being programmed to recognize fruit size and color, and to ‘see’ fruit and pick it as the machine moves down the row.


The current configuration does put fruiting wood in a position that fruit may develop a bar bruise or cut from the fruit passing along the fruiting wood on its way through the harvester. The most proficient cure to this horticulturally preventable problem is to fasten the fruiting limbs longer than 9 inches in a manner they are not taken into the harvester. Short fruiting structures will likely minimize the problem.

Vertical or Angle canopy? For best results for the current version of technology, there should be nothing between the ‘end effector’ and the fruit. No branches, twigs, posts or wires should be directly between the picker and the fruit. Wood, other than a short bourse shoot that comes directly into the row is a particular problem. The very organized (all fruiting wood fastened to wires, or is very short bourse shoots/spurs) canopies, either vertical or angle should be very productive / compatible with robotic harvesting. The very narrow canopy provides much more uniform solar radiation (heat on fruit), light distribution and fruit distribution. A much higher portion of fruit can be ready for first pick in well balanced, ultra-narrow fruiting walls.

Random training, where only the more vertical leader is fastened to wire is more challenging, for relatively few fruiting limbs are fastened to wire. Mechanical pruning appears to be a very quick and consistent method of minimizing problem wood. In the longer perspective, if the harvester can fit into 10 foot centers (+ or – 2 feet to be able to efficiently work in rows with variable width of 8 to 12 feet) then vertical systems should be strongly considered to utilize the mechanical pruning, thinning and harvesting technologies.

Variety selection. Some varieties, especially those harvested late in the year or attractive to birds, can have a number of rotten apples that will likely get picked and smatter their contents inside the machine and successive fruit entering the machine. Inoculating fruit with disease is likely the smaller problem, for the bigger issue may be the sticky apple film’s attractiveness to abrasive materials that could abrade fruit following the messy one, but also mechanical/physical issues related to ‘apple goo’ mucking up the equipment down stream such as in the bin filler.

Future Synergies

If the harvester with bin filling can function in the older 10 foot center blocks (8 to 12 foot rows) the opportunity to nearly fully automate the apple-pear industries are at hand. Planting systems utilizing multi leader trees spaced at ~22 inches (plus or minus 4 inches) per leader with 2 to 3leaders per stump become theoretically possible. 2 leaders per tree at 18 inches between leaders will place stumps 3 feet apart, 3 leaders at 26 inches is 6.5 feet between stumps. As the distance between leaders increase, and the number of leaders increase per stump, the ‘likely to succeed factor DECREASES. It is important to assure rapid and UNIFORM production canopies are easy and efficient to establish. As the number of leaders/stump and the distance between leaders increase, the amount of management required to grow consistency of leader height and fruiting structures increases. ‘Management’ is manipulating strong leaders into an angle to reduce vigor or defruiting the weak leaders until they ‘catch up’ to the strong leader.

If the harvester requires a wider foot print, then angle canopies ability to increase the bearing surface/yield may become important. A 12 foot angle canopy will increase the number of leaders spaced at 22 inches by 66% per acre (from 2376 in a 10 foot vertical row to 3960 in a 12 foot angle canopy row) The productive capacity of each leader may be reduced to keep the fruit on the outside of the wire, and fruiting structures pointed into the interior of the V, may not be harvestable by machine. Also, the interior of the V will need to be manually pruned / or thinned.

Conceptually, some varieties may be able to be green fruit thinned by the robotic harvester. The narrower the canopy, the better the fruit will be viewed / counted for thinning.

Evolutionary improvements will advance the robotic technology and also the orchard systems and their management. First attempts will be followed by increasingly successful second and third implementations. The WSU Tree Fruit Extension Team will work to identify the most critical traits of orchard systems and their management along with varieties that present the fewest (or most) challenges to harvest robotically.

As a foot note, the tools available to mechanically prune and adjust the pruning has evolved considerably in the last 5 years. The ability to plant very straight rows (which will enhance machine performance) has improved significantly in the last 5 years.

tom_auvilWritten by: Tom Auvil, Program Leader,
Washington Tree Fruit Research Commission,
Wenatchee, WA.

Washington State University