ROAM'ing Towards Success in Mapping the Eradication of Invasive Pests in the US
In a pivotal meeting held in mid-December 2015, GIS staff and Plant Health Officers from the USDA's Asian Longhorned Beetle (ALB) program demoed ROAM (a field mapping application based on QGIS) to IT and ALB program managers; successfully making the business case that it was finally time to approve a well-tested piece of Open Source GIS software for use in the agency. In a largely unprecedented and uncommonly swift decision, senior management approved the ALB Program to move forward with a ROAM deployment for all ALB field offices within a week's time, and the infamously slow IT Security review process was slated for completion and final approval by late January 2016.
The Asian Longhorned Beetle is a destructive invasive pest that has infested several states in the north-eastern United States over the last 20 years; and put several economically important hardwood tree species at risk of decimation - particularly maples. It has been successfully eradicated from Chicago, Boston and New Jersey, but the battle is still under way in south-western Ohio, central Massachusetts and on Long Island and parts of New York City.
Prior to spring of 2015, the ALB Program was using a legacy, non-spatial PDA application for field tree survey and removal data collection; as well as a legacy back-office data processing application that restructured the PDA text files into records compatible with an Oracle database. In late spring of 2015, due to security concerns after a number of non-related IT security breaches at Federal data centers, many applications running older versions of Java and Oracle that could not be successfully upgraded in a short amount of time were taken off-line. This included the ALB Program's database and web-based survey data management application. At the time, a contractor was in the process of developing a new mobile field data collection application for all of the pest programs in PPQ that would interface with a newer Oracle enterprise data repository. ALB staff continued to use the legacy PDA application to collect survey data in the field - but lacked an automated means to process it, or a database to store it in. Data files were stored on the local network file servers in each office, waiting for the new IT solution to be ready.
Over the remainder of 2015, it became clear that the mobile application would be very limited in its spatial capabilities, compared to most modern GIS software tools. The contractor did not have a background with the type of geospatial development expertise needed to create that functionality. After months of testing, ALB program management determined that while the new mobile application could be adequate for other agency pest programs with less intensive geospatial requirements, it would not meet the ALB program’s needs. ALB staff must survey both urban neighborhoods and dense rural woodlots, occasionally needing to climb all the host trees on the property to inspect for ALB damage or presence. A robust field mapping application with modern GIS functionality and the flexibility to incorporate local government spatial data on an ad-hoc basis in each of the infested areas was required.
In the meantime, ALB staff had been searching for a 'Plan B' in the event that the contractor-developed solution did not pan out; and they found one in ROAM - a field application based on QGIS. The major hurdle was that Open Source Software was considered “highly suspect” by agency IT Security and not allowed on the network. Fortunately, staff discovered that ROAM does not require a traditional Windows installation process and can be successfully run from a USB memory stick on a tablet out in the field - without actually installing the software on the field computer. ALB staff in one of the field offices created a prototype ROAM ‘exploratory’ project and tested it for several months in the field; and it proved to work extremely well.
Efforts to modify the original ROAM data collection project and standardize the file geodatabase structure are currently under way at the time of this paper proposal submission, and an initial deployment is planned in all 4 ALB offices by the end of February, 2016. ROAM's extreme ease of use, as well as ease of set-up has facilitated such a rapid deployment schedule for the initial ALB Survey data collection module. Over the course of spring and summer 2016, additional modules to collect data for Callbacks, Removals, Treatment and Regulatory Permits will be created and deployed. This presentation will outline the initial ROAM implementation process and discuss what worked, what didn't, and how solutions for success were devised.