Updating PMapp, a smartphone application for assessing powdery mildew

Abstract

Diseases, pests and disorders can compromise the yield and quality of grape harvests. A smartphone application, PMapp, was developed in project UA1202 to facilitate visual estimation of powdery mildew severity on bunches. In response to requests from users, the GPS capability was improved to allow more accurate recording of latitude and longitude. Also in response to demand, the assessment component of PMapp was expanded to allow the user to record in-field assessments of other conditions and to customise the assessment screens, in a new app, Grape Assess. The resulting apps, for Apple and Android systems, were released in June 2018 following evaluation by the Project Steering Group.

Summary

Diseases, insect and vertebrate pests, and disorders can compromise the yield and quality of grape harvests and have the potential to cause considerable loss to the Australian grape and wine sector. Powdery mildew, downy mildew and bunch rot together have been estimated to cost the sector $191 million per annum. Many Australian wineries set thresholds for contamination of grape consignments caused by diseases and pests, commonly 3-5%, above which consignments may be downgraded or rejected. In the absence of widely applicable objective measures, such assessments are typically based on visual inspection in the vineyard and/or at the winery. An application for smartphones and tablets, PMapp, and a supporting website were developed as part of project UA1202, Objective measures for powdery mildew, with the aim of improving the accuracy and reliability of visual assessment of powdery mildew on bunches. These resources have been available since the 2015-2016 growing season.

Feedback from users of PMapp indicated that the tool was useful and generally well-received, but that the GPS (global positioning system which records latitude and longitude) was insufficiently accurate to provide the level of detail required by some assessors. Furthermore, the app was being used to assess diseases and disorders other than powdery mildew. The research reported in this document was undertaken to (i) improve the accuracy of the GPS component of PMapp and (ii) create a new app with expanded capability to allow users to record assessments of conditions other than powdery mildew.

The settings of the upgraded PMapp and the new app, Grape Assess, are such that the GPS offers accuracy of approximately 10 m, although this can depend on location. The additional features of Grape Assess are based on discussions with members of the Project Steering Group, the recommendations contained in the final report for project UA1202 and the quantum of funding available. Grape Assess provides the four components of PMapp, namely; (i) assessment data entry, (ii) image browser, (iii) self-calibration test for visual assessment and (iv) a diagrammatic key with 2% increments in the range 2-12% severity. The computer-generated images used in components ii-iv were designed to reflect the distribution of powdery mildew on grape bunches. The user is able to select from a built-in list of assessments and/or to customise assessments according to the condition and vine part being assessed, the severity categories required and the background colour for the buttons on the assessment entry screen. Grape Assess is linked to the online powdery mildew assessment resource at pmassessment.com.au. The detailed guidelines for use have been updated to reflect the new capabilities; in particular, a section on visualising the vineyard assessment has been added and codes prepared using the free computer software, R, made available on request. The updated PMapp and Grape Assess were made available globally as free downloads for Apple and Android devices in June 2018.

Outcomes have been communicated through the Project Steering Group, and users of PMapp will be prompted to update this app and to consider downloading Grape Assess. A presentation to industry is planned in the coming months, as are publications in industry and scientific journals; the short time-frame of the project limited such communications before submission of this report. PMapp and Grape Assess are or will be used in teaching viticulture and oenology students at the University of Adelaide, the University of Tasmania and Charles Sturt University.

Wine Australia and the School of Agriculture, Food and Wine of the University of Adelaide provided financial and in-kind support. Collaborators include RW Emmett Horticultural Pathology Research, the Tasmanian Institute of Agriculture, Accolade Wines and Lemur Software. This project benefited from an engaged and enthusiastic Project Steering Group that comprised personnel of large and small wine sector companies, research organisations and Wine Australia.