Parks Observer Server and Web Application


Vladimir Bukhin

Oral Defence Date: 

Wednesday, February 27, 2013 - 15:10


TH 434


3:10 PM


Professors Marguerite Murphy and Dragutin Petkovic


The SFSU Observer Project consists of client devices sending observational data of different varieties to servers and social networks for aggregate viewing and management purposes. This report describes the design, implementation, verification, and evaluation of a configurable web application and API to host and manage variable observer themes. Constrained data storage and retrieval is done using an indexed and configured Mongo DBMS. The server software is written in NodeJS with the Express framework while the browser side application is written in Javascript with the AngularJS framework while using Twitter's Bootstrap CSS style library. A usability study showed that label-naming in certain content and out of context instructions in the web application caused confusion for participants, and the user interface was redesigned to address this issue. The contributions of this work include a configured Heroku cloud environment with a MongoLabs configuration, a flexibly programmed prototype of the web application that allows the observation theme and data fields to be configured, a content delivery service that allows access to binary files stored on the MongoLabs. GridFS, a public API definition to support development of mobile Apps for submitting applications, a user interface design to both display and allow management of observation data, and a survey-based usability study.

Vladimir Bukhin

Field data observation, NodeJS, Heroku Deployment, MongoDB, Twitter Bootstrap, web application, API, content delivery server, AngularJS