Review
Soft Grippers for Automatic Crop Harvesting: A Review
Eduardo Navas * , Roemi Fernández * , Delia Sepúlveda , Manuel Armada and Pablo Gonzalez-de-Santos
Citation: Navas, E.; Fernández, R.;
Sepúlveda, D.; Armada, M.;
Gonzalez-de-Santos, P. Soft Grippers
for Automatic Crop Harvesting: A
Review. Sensors 2021, 21, 2689.
https://doi.org/10.3390/s21082689
Academic Editor: Dionysis Bochtis
Received: 10 March 2021
Accepted: 9 April 2021
Published: 11 April 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright:
c
2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Centre for Automation and Robotics, UPM-CSIC, Carretera CAMPO-REAL Km 0.2, Arganda del Rey,
28500 Madrid, Spain; delia.sepulveda@csic.es (D.S.); manuel.armada@csic.es (M.A.);
pablo.gonzalez@car.upm-csic.es (P.G.-d.-S.)
* Correspondence: eduardo.navas@csic.es (E.N.); roemi.fernandez@car.upm-csic.es (R.F.)
Abstract:
Agriculture 4.0 is transforming farming livelihoods thanks to the development and adop-
tion of technologies such as artificial intelligence, the Internet of Things and robotics, traditionally
used in other productive sectors. Soft robotics and soft grippers in particular are promising ap-
proaches to lead to new solutions in this field due to the need to meet hygiene and manipulation
requirements in unstructured environments and in operation with delicate products. This review
aims to provide an in-depth look at soft end-effectors for agricultural applications, with a special
emphasis on robotic harvesting. To that end, the current state of automatic picking tasks for several
crops is analysed, identifying which of them lack automatic solutions, and which methods are
commonly used based on the botanical characteristics of the fruits. The latest advances in the design
and implementation of soft grippers are also presented and discussed, studying the properties of
their materials, their manufacturing processes, the gripping technologies and the proposed control
methods. Finally, the challenges that have to be overcome to boost its definitive implementation in
the real world are highlighted. Therefore, this review intends to serve as a guide for those researchers
working in the field of soft robotics for Agriculture 4.0, and more specifically, in the design of soft
grippers for fruit harvesting robots.
Keywords: soft robotics; agriculture 4.0; soft grippers; end-effectors; review; harvesting process
1. Introduction
In the last decade, the agricultural sector has undergone a deep transformation to cope
with the growing demand for food [
1
–
3
]. Among the main tasks in agricultural processes,
those that involve the manipulation of fruits and vegetables continue to be one of the most
time consuming and labour intensive, resulting in low efficiency and limited competitive-
ness. This situation is exacerbated by the labour shortages of seasonal workers unable to
travel between regions, leading to the accumulation of fresh products and impressive food
losses. For these reasons, a great research effort is underway to automate these manual
operations, as in the case of selective harvesting, combining multidisciplinary fields such
as biological science, control engineering, robotics and artificial intelligence. Special em-
phasis is being placed on topics such as the modification of plant peduncles [
4
], which
could simplify the harvesting process [
5
]; machine vision and detection
systems [6–10]
;
decision-making architectures [
11
–
13
]; autonomous navigation [
14
–
16
]; and dexterous
manipulation [17,18]
. Another critical topic, often underestimated, is that related to the
design of the systems attached to the tip of robotic manipulators and that are in direct
contact with the fruit, known as grippers or end-effectors.
In manual harvesting, humans use their hands to move different elements of plants,
grasp the fruits and detach them, either directly or with the help of a tool. The kinematics
of human hands, the deformability of the skin and muscle, and their sense of touch give
us efficient grasping abilities. Attempts to emulate human skills during harvesting have
resulted in numerous mechanical end-effectors that can be classified according to their
numbers of fingers into two major groups: multi-fingered and parallel grippers [19].
Sensors 2021, 21, 2689. https://doi.org/10.3390/s21082689 https://www.mdpi.com/journal/sensors
