Using a mobile phone Short Messaging Service (SMS) for irrigation scheduling in Australia – Farmers’ participation and utility evaluation
Highlights
► Simplicity of this DSS interactions reverses the trend in irrigation DSS design. ► Eighty percentage of irrigators in this trial found this system useful. ► Interviews show simplicity, flexibility and intrusiveness of interaction are key. ► SMS offer a simple, bi-directional irrigation DSS interface for complex models. ► SMS has close-to universal deployment possibilities among farmers.
Introduction
Irrigation scheduling Decision Support Systems (DSS) have experienced poor uptake amongst irrigators in Australia despite much investment and well publicised objective evidence that they can increase water use efficiency. For example, WaterSense, a sugarcane irrigation scheduling DSS with proven ability to increase water use efficiency (Inman-Bamber et al., 2005) has fewer than a hundred users among a potential pool of thousands. Australia-wide, a 2005 survey of irrigation DSS found that 21 are in operation but most have only a dozen or fewer users (Inman-Bamber and Attard, 2005).
A survey by Olivier and Singels (2004) of the reasons for not adopting scientific irrigation scheduling techniques by South African irrigators identified two main barriers to adoption. The first was the complexity of use and hence the difficulty of applying them to farm practice and the second was whether their use would actually translate into benefits. Much work in Australia on barriers to DSS adoption (Carberry, 2001, McCown et al., 2006) notes a ‘gap’ between scientific and industry approaches to scheduling that supports the South Africa experience.
Analysing a typical irrigation scheduling tool such as ‘WaterSense’ (Inman-Bamber et al., 2005) we see that it requires Internet access, the entry of much data and a long wait time (40 min) for results to be generated. WaterTrack Rapid (Watertrack, 2006), another recent Australian DSS, designed to minimise user effort through limiting data reporting,1 still requires irrigators to access the Internet from a computer, enter in data and then run the system for a response. Given the known rate of adoption of all DSS in Australia, is extremely low (Inman-Bamber and Attard, 2005) and knowing the well understood barriers to adoption, it is clear that in the case of WaterSense, the benefits to its use do not outweigh its difficulty of use. In the case of WaterTrack Rapid, its use is still not sufficiently easy to result in widespread adoption.
Irrigators who do use some form of decision support do so only in conjunction with many other data sources. Pre-trial interviews with our test group of irrigators showed that all irrigators, including those who had soil moisture probes installed, used vine observations, weather observations and experience to help them irrigate. Many used soil wetting pattern observations and shovel soil testing and a few individuals also used less common data sources, such as infra-red crop images or weather warnings from other locations as a forecast to help them. For many of these data sources there is no objective, deterministic method that can be used to generate irrigation volumes. Invariably, irrigators must rely on heuristics and past experience to do this. Thus DSS such as SOAK (Costigan, 2008), a farm water management software package that tries to coalesce all the irrigation data sources that its designers believe are relevant to irrigation scheduling, must inevitably fall short of providing ‘all the answers’ for many irrigators both for their inability to include the whole, wide, range of data sources currently in use and their inability to measure the non-quantifiable sources.2 Since the announcement of the development of SOAK and a prize it won in 2008, there have been no further references to it publicly in the irrigation industry.
The cellular mobile Short Messaging Service (SMS) is increasingly used in many contexts to simply and quickly deliver and gather data from people with mobile phones. One recent example of its’ use is by diabetic clinics in managing remote patients’ blood sugar levels (Hanauer et al., 2009). In irrigation, SMS has been used to promote the understanding of how a flexible, water budget-based, irrigation schedule can save water and increase productivity over a fixed schedule (Singels and Smith, 2006). In this South African trial messages were sent weekly to 5 irrigators telling them to “stop-”, “start-”, “continue-” or “do not” irrigate based on a crop growth model using estimated irrigations and measured rainfall. The study concluded that by communicating the model outputs to irrigators in the simple SMS form, water use efficiency was increased by 48%. The authors found that weekly communication was required to assure the participating irrigators that the system was still functioning. They also concluded that it could be advantageous to obtain measured irrigation volumes from the participating irrigators to improve model accuracy (Singels and Smith, 2006).
In addition to its ease of use, the low deployment cost of SMS and its ubiquity of use by many people, even the poorest, in both the developed and developing world means it is a technology that helps bridge the digital divide rather than widen it, thus modified forms of systems used in the developed world may be more easily modified for use in the developing world than Smart Phone or personal computer-based technologies.
This paper describes the use of high-end IT systems for irrigation scheduling and the response of irrigators to an SMS-based, textual, DSS interface for irrigation scheduling. It also reports the utility of such an approach for irrigation scheduling as determined both by system data and interviews with end users. Methods detailing technical aspects of the system, the trial location and participants, interactions with participants, the DSS calculations and the DSS operation in general are given. Results of system performance, irrigation applications and system cost to irrigators are provided from measured data. Results of irrigators’ understanding of the system and perceived system utility are provided from interview data. Discussion of user participation and user utility follows and finally a conclusion is given.
The motivation for this research is the desire to improve the utility of DSS to irrigators in order to increase the proportion of them using such tools. Increased use will not only lead to benefits to the individual but also to the irrigation industry as a whole as such DSS also easily function as data collectors for aggregate water use statistics.
Section snippets
System description, calculations and operation
The DSS used in this trial was called IrriSatSMS, which used satellite derived crop coefficients in a daily water balance approach. The mechanisms of the water balance calculations and crop coefficient generation and use are described in detail in Hornbuckle et al. (2009). An overview of the Decision Support System’s (DSS) architecture, as well as the methods of communication used between components, is shown below in Fig. 1.
The central server hosted the DSS calculation code (Microsoft C#.NET
Results
The service began in September 2008 with 23 irrigators and this number increased through the early part of the season. By November 2008, 54 irrigators were enrolled, by December 2008, 67 and from January 2009 onwards, 72.
Participation
For participants in this trial, cost was virtually non-existent and the labour effort required was very low. Receiving and sending messages throughout the season did not perturb the majority of irrigators, no irrigators that discontinued with the trial cited reading or sending messages as the only issue leading to them not using the DSS. It appears that the effort required for effective participation did not exclude any irrigators from using the DSS. The complexities of ET-based water balance
References (13)
- Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration – Guidelines for Computing Crop Water...
Can science rigour and industry relevance both be achievable in participatory action research?
Agricultural Science
(2001)- Costigan, G., 2008. Soaked in success. Available from: <http://archive.uninews.unimelb.edu.au/news/5336/index.html>...
- et al.
Computerized Automated Reminder Diabetes System (CARDS): E-mail and SMS cell phone text messaging reminders to support diabetes management
Diabetes Technologies & Therapeutics
(2009) - Hornbuckle, J.W., Car, N.J., Christen, E.W., Stein, T.M., Williamson, B., 2009. IrriSatSMS – Irrigation Water...
- Inman-Bamber, G., Attard, S., 2005. Inventory of Australian Software Tools for on Farm Water Management,...
Cited by (52)
Decision-support systems for water management
2022, Journal of HydrologyDecision support systems and models for aiding irrigation and nutrient management of vegetable crops
2020, Agricultural Water ManagementA survey on intelligent agents and multi-agents for irrigation scheduling
2020, Computers and Electronics in AgricultureEvaluation of a new irrigation decision support system in improving cotton yield and water productivity in an arid climate
2020, Agricultural Water Management