1Lincoln Technology, Lincoln Ventures Ltd,
Private Bag 3062, Hamilton, New Zealand
2New Zealand Centre for Precision Agriculture,
Massey University, Palmerston North, New Zealand
Ph 0064 7 858 4852, Fx 0064 7 858 4840, firstname.lastname@example.org
Horticultural enterprises are well suited to site specific data techniques and information management systems. These techniques include using GIS database structures, handheld data capture devices, remote sensing equipment, database management at different levels in the supply chain. They can be used to optimise physical and financial aspects of horticultural production systems along with provision of reliable audit for product security. While there are exciting opportunities to further progress, the consideration of horticultural production in the context of its supply chain offers some challenging requirements for these technologies, not the least of which are questions such as “how will it work in practice”. This paper reports on progress to date in New Zealand and explores the implications for researchers, developers and users of the various technologies.
Information systems, management, horticulture, product tracking, supply chain
In order for Southern Hemisphere fruit production to compete in the global marketplace it needs to be become more sophisticated. There is a critical need for the producer to be more intimately linked with their consumers. Producers need to know what the critical determinants of value are in the consumers mind, and endeavour to produce fruit which meets these as well as the constraints of the supply chain.
The determination of quality at present is primarily based on firmness and visual attributes. The development of new technologies means the requirements will expand to include various flavour and storage capability attributes. Current market feedback indicates that one major driver is the increasing need to provide a consistent quality product. The variability in quality on out-turn is one of the major complications for the supply chain. This can be attributed to a range of postharvest factors (storage time, location on ship, pack type, etc). On top of these system induced variations there are a large range of orchard factors which produce variation in the initial fruit quality as well as its inherent storage and eating quality (Figure 1). Little is currently made of this orchard induced variability.
There is a need for improved segregation and traceability of product through the supply chain to extract the maximum revenue from increasingly sophisticated future market segments. Production and postharvest techniques and inventory management procedures will need to become more sophisticated to address this challenge. If orchardists are to be rewarded for optimising their quality and produce a more uniform product, then systems are required to facilitate this.
Current situation and opportunities
The current supply chain only provides minimal feedback to producers on the critical determinants of value. The more sophisticated this feedback can be, and the closer the consumers are brought to the producer, the more opportunities growers have to produce optimised fruit quality. The enabling technology for data capture and the information management systems to facilitate these developments do not currently exist at a level which is appropriate. Real-time monitoring, databasing, analytical capabilities and information visualisation techniques offer the opportunity to improve information flows and empower businesses within the supply chain by providing tools to monitor and respond to consumer requirements. However, few of these technologies have been applied despite the fact that biological products have a large inherent quality variability which makes them a prime opportunity to implement complex analysis and methodologies. Variability in quality attributes such as size, colour, shape, flavour, sweetness and firmness all detract from the value proposition at the point of sale. Improved (more detailed) information of on-orchard variation of these factors will help to manage fruit quality (Whitney et al., 1999). Reliable knowledge of product origin and safety is also becoming a market demand and requires systems for cost effective product tracking. Profitability is a key concerns for participants in the supply chain which is a function of product value, quantity sold and production costs. Therefore a combination of various technologies which provide the tools to satisfy product quality, safety and profitability requirements are necessary for maintaining supply chains and may form part of a precision horticultural approach to management.
Orchardists have a range of management considerations to make. These will impact in various ways on the quality of fruit at the packhouse, the coolstore and the market. Some of the relationships are known and some are unknown. The subtle interactions specific to an individual orchard are not known. Capturing information for a small area of the orchard (a group of trees/vines) and then tracking this fruit through the packhouse, and quality audits, will facilitate this.
In addition there will be information that can be collated through such a system to regional or a national level that can provide other data and models (Figure 1).
Figure 1 Fruit quality and orchard operation information structure
Critical to the full implementation of a precision horticultural system, then, is a Global Information System. In the broadest terms this means a system which enables a free flow of relevant information to be available to all the relevant participants in the supply chain (Figure 1). The system needs to, for example, provide retailers and consumers information on the grower and growing details at the supermarket, or feedback in-market quality assessments to growers, who can trace this back to areas within their orchard. The system needs to integrate horizontally (along the supply chain) as well as vertically. This vertical integration allows further enhancement of information through the collation of regional and national trends. It also offers opportunity to develop models that are more robust than those derived from individual participants in the supply chain. There are no current supply chains which can offer the level of sophistication required to take full advantage of the opportunities offered by precision horticulture.
A major data structure gap is the linkage between of packhouse data and specific sites on the orchard. Linking packout data with local production sites will potentially provide growers with a quantum increase in the level of sophistication of the data they can access for management decisions, and will also provide the tools to turn every orchard in New Zealand into a micro-research orchard. At present, research orchards develop specific information on optimal orchard practices. This information is difficult to apply at orchard, regional and national level.
While the overall concept of improving information management systems for the supply chain is a large and complex issue it will only be solved by addressing individual issues. We have made progress on addressing some particular issues which we believe are critical to take advantage of the opportunities outlined above.
Apple orchards are laid out in a series of small blocks (0.5 - 2.0Ha). They often have a range of varieties grown in each block and trees of differing ages. Current data from the packhouse comes back to the grower as an orchard, or at best block, average. In this study the harvest and packing of Braeburn and Pink Lady varieties were recorded and analysed at the sub-block level.
Apples were picked into 400 kg bins from a group of trees and transported to the packhouse for grading and packing into 18kg cartons. The location and identity of each bin was recorded during picking using hand held electronic ID tag technology. The data was then loaded directly onto the database at the end of each picking day. The size and colour of each apple was recorded at the packhouse at the time of grading and linked to individual bins. This data was then linked with a group of trees in the orchard where the bin came from and mapped accordingly (Figure 2). Once the link between the bin and the graded apple is established it becomes possible to track post-packing data such as storage and quality characteristics back to orchard site via the bar code on individual cartons.
Figure 2 Spatial variation in colour across a block of apples (darker (red) region indicates high proportion of high colour fruit)
The future development of the New Zealand wine industry depends on growth in export sales. To maintain this growth wine style must be maintained between seasons. Fruit is predominantly machine harvested and as a result the wine quality represents the average fruit composition for a block. In recent years the variation in this quality about the mean has been recognised as a key determinant of wine quality. As a result, understanding the variation in fruit composition within the vineyard and the impact that management, soil type and other environmental factors can have on that variation is important.
One of our current research projects is targeted at developing new tools for viticulture management. A range of sophisticated modern technologies is being evaluated for their ability to provide real-time information on canopy and berry attributes to assist with husbandry and harvesting operations. They include laser scanning and ranging, visible and NIR imagery, ultrasonics, and the measurement of light interception. Laser scanning is described here.
An LMS 220, a laser scanning and ranging unit manufactured by SICK AG, Germany, is being evaluated for measuring grape canopy architecture (row volume and porosity). Knowledge of row volume and porosity allows spray applications to be adjusted to minimize agrichemical use and off-target losses. Canopy architecture also reflects canopy health and vigour which can be related to grape and wine quality. The objective is to determine the viability of integrating LMS measurements and Geographical Information in real time to record canopy characteristics across vineyards to manage spraying and harvesting operations.
For the initial evaluation, the LMS was attached to a vertical pole mounted on the back of a ute which could be driven between the rows of the vineyard (see Figure 3). The ute was stopped at each measuring point and several vertical scans normal to the row were recorded from two heights above ground level, one at 600mm (just below the lowest grapes) and one at about 1800mm (near the top of the canopy). The repeatability of the data, the effect of LMS height relative to the canopy, and possible measures of canopy porosity are currently being evaluated. These electronic signatures of canopy architecture will be calibrated against physical canopy measurements and compared with fruit quality measurements determine if canopy scanning can be used for fruit and hence wine quality prediction. The data will analysed spatially to assess the variation across the study block.
Figure 3 Testing the laser scanning unit for canopy measurement in grapes
Significant variation in fruit quality has been shown across relatively small blocks (ca. 0.5 ha) of apples and grapes. This data can be related to profitability (Praat et al., 2001). The strategy of integrating GIS databases to better utilise existing data and product tracking techniques can address broad management issues and product origin concerns. However the full value of the data will only be appreciated when tools are further developed to enable more statistically robust and more complex interactions that characterise the relationships between the variation in orchard characteristics and the quality outputs to be dealt with. Visualisation of fruit quality on-orchard increased the awareness of growers to consumer preference. Maps which show growers the profile of attributes which consumers prefer for example colour or dry matter can also enabling growers to start to manage and optimise these characteristics as well as the traditional attributes of size and quantity.
Issues to consider
The system also has potential to offer other advantages to growers, such as picker payment incentives (as the reject rate for individual bins can be measured) and monitoring coolchain performance after harvest. Integration of other data capture techniques such canopy scanning, electromagnetic survey, nutrient mapping and infra-red sensing with product quality maps can help to identify factors critical to product quality.
An issue of major importance is that the integrity of the fruit quality data is determined by the degree to which fruit from one bin is mixed with another during the grading and packing process. The degree of mixing will vary with the system and tends to be specific to individual packhouse operations. Consideration needs to be given to tracking potentially mixed and complete bin samples in the final packed unit. This issue may limit the reliability of the tracking system and for bin separation, may feature in future packhouse design.
The work to date has also identified the constraints to the implementation of practical systems based on these technologies. Current GIS tools are not necessarily cost effective for individual orchardists, and the time taken to enter the data and produce the maps was complex and time consuming. Much of the useful data had to be collected using a combination of manual and reasonably expensive electronic techniques that would be too expensive for a practical application. So while the potential for this technology has been clearly demonstrated, the actual value of the information to growers does not currently justify the costs involved in collecting and presenting it. However the requirement for reliable product tracking capability may speed the development of these systems as consumers become more concerned about product security.
Experience with yield mapping of broad acre crops suggests that the benefits of applying site specific farming techniques are not always obvious at the outset but that the costs are offset by improved quality of management information which is gained and can be used in decision making (Yule et al., 2001).
This research is starting to demonstrate the power of this technology in the horticultural tree crop industries. By providing the appropriate technologies for data capture, in time it will be possible to have a seamless information system for horticultural products that will track fruit from the tree to the final market. This will enable delivery of quality information from the market back to the producer. By detailing this data to small areas within the orchard, through the application of new spatial data interpretation techniques, the data can be transformed into powerful knowledge tools. This will enable growers to target their production to the market related issues which impact most significantly on their bottom line. The technology could also potentially open up marketing opportunities, as increased knowledge of the production system furnishes the grower with the ability to segregate and reliably deliver fruit with specific quality attributes.
The major issue for practical application is the ability to cost effectively capture the data and cheaply produce the visualisation maps. This is the future for applied information management systems.
We would like to acknowledge the financial support for this work from Electronic ID Tag Systems, Ltd; Technology New Zealand and the Foundation for Research, Science and Technology. This project was made possible by the patience and cooperation of Martin Reid, Richard Glen and Montana Wines Ltd for which we are also very grateful.
Praat, J-P., Bollen A.F., Yule, I.J. (2001). Product tracking for profit In: Precision Tools for Improving Land Management, Proceedings of the 14th Annual Workshop, Fertiliser and Lime Research Centre. In Press
Yule I.J., Praat J., Craighead M., Austin J. 2001: Opportunities from Precision Agriculture. In: Proceeding of the NZ Maize Conference 2000; Agronomy Society of NZ Special Publication In Press
Righetti, T. L., Halbleib, M. D. (2000) Pursuing precision horticulture with the internet and a spreadsheet. In: HortTechnology. 10 (3):458-467
Whitney, J.D., Miller, W.M., Wheaton, T.A., Salyani, M., Schueller, J.K., (1999) Precision farming applications in Florida citrus. In: Applied Engineering in Agriculture. 15(5):399-403