Agriculture Reference
In-Depth Information
cutting, grafting and transplanting, and the final phase of sorting and packing the
harvested produce can be mechanised. These tasks do not require much human
intelligence and/or fast and accurate eye-hand coordination. Currently machines
are largely based on principles of industrial automation consisting of mechanical
solutions with only a limited amount of sensors and 'intelligence' being used. Crop
maintenance and harvesting operations do rely on human intelligence and ability
and are more difficult to automate. It is anticipated that in next 10 years will see
the advent machines that will be based on the principles of mechatronics and robot-
ics, combining smart mechanical design with sensors and “artificial intelligence”
achieving the fast and accurate artificial eye-hand coordination needed for these
difficult tasks. The slow progress in developing robotic harvesting is due largely
to uncertainty in the working environment of the robot resulting from biological
variability and the types of growing systems used. Progress requires innovations
in robot technology, growing systems and plant breeding which collectively reduce
biological variability and simplifies the tasks involved. In particular, the genetic
profile of each cultivar will need to be known in minute detail. As a consequence the
functions and processes controlling the abilities of each gene and, more importantly,
perhaps their interactions will be matched with known responses to environmental
factors. One consequence of this new level of sophisticated husbandry is to ask
whether horticultural cropping will remain field-based?. In the past decade, soft
fruit has moved under protection and the tree fruits are moving in that direction.
Considerable use is made already of automation and robotic technologies in the
post-harvest and storage phases, particularly in the three major steps of grading,
sorting and packing (Feng et al. 2010 ). Assembly lines emerged in horticultural
packinghouses early in the 1900s. Major advances in the removal of manual labour
have however, only arrived with the advent of electronic and optical systems which
have operating speeds comparable with human eye decision making. Machines need
to detect the colour, defects, size, shape, volume and density of produce. That can
be achieved with techniques such as optical machine vision, near-infrared radiation,
X-ray, and acoustic responses allowing 3-D machine evaluation of horticultural
produce. Unlike industrial products, horticultural produce is variable in size, shape,
volume, density and orientation. Automation and robotics must be able to cope with
these factors. Frequently the unconstructed and tough environment in a packing-
house or the field at harvest challenges the operation and safety of the machines.
Downtime resulting from machinery malfunctions must be minimised since loss
of capacity in the packing house can have penal implications for the efficiency of
the logistical systems designed to deliver produce into retailers' premises on a “just
in time” basis. Machinery manufacturers are employing wi-fi technology which
reports on the operating efficiency of equipment directly to control centres. These
technologies issue alerts and fault diagnostic reports to mobile manufacturers' staff
who are then able to prevent or minimise machinery downtime. In large scale post-
harvest processing of crops such as Citrus one major problem is the detection of
visual defects and the presence of undesirable matter (Lopez-Garcia et al. 2010 ).
Species and cultivars of Citrus present a high rate of unpredictability in texture and
colour. Detection of defects in real-time based on random colour textures, using
Search WWH ::




Custom Search