Scansistant

Nowadays a lot of students have smart-phones that can be put to use for mobile learning. The TU Delft Library has come up with an idea for this which also incorporates the use of a modern technology called Near Field Communication. It is already available on the market but immature so experience has to be gained in combination with mobile learning to see if it is worth to pursue further developments on this area. The TU Delft Library has the idea of using a mobile application to scan machinery on the campus as a way to get instructional videos and information to perform a feasibility study and to gain experience on the aforementioned area of interest. From the wishes of the client and our own acquired computer science skills the requirements came forth and as a start of the software development. For the requirements of the final delivered prototype application we carefully looked if the features could help to perform the feasibility study and are able to answer the questions for the this study. From these requirements we conclude that more than a simple Android application has to be built: a central server to house all video and data was also required and a way to provide communication between this server and android client application. For this we designed an API so that many clients can get information from the server. The software subsystems consist of a back-end, front-end and client application. They are developed in Java, building upon the Android SDK and in PHP, building upon the Zend Framework 2. For the NFC technology libraries are available so that tags could be successfully scanned after which the machine data could be downloaded. For development of all subsystems we used Eclipse and associated plugins to enable for Android and ZF2 development. The design of the subsystems was done in parallel and on-demand so that features required by the android application that needed an API method that was not yet there was immediately implemented when needed. It started with a global database design after which we could both focus individually on the platform specific details. The user interface design was important because that is the level at which the feasibility study will be performed when the end-user gets to use the application. The design focus was on smart-phones, but because tablets are widely used also we made it compatible making use of its bigger screen. Because of an agile and feature driven development methodology we use exploratory testing for the largest part of the project. After we had a stable system we started writing tests because we now had gained enough knowledge of the development platform to efficiently write tests. At the end we managed to have tested the Android application by unit testing the models and the API was tested by a controller testcase to test the responses of critical API methods. For the scope of our project we did not need to perform the feasibility study ourselves, but we did ran some user evaluation trials to get a first reaction on the application before delivering a final prototype. The feedback that we got from these user evaluations allowed us to increase the chances at success because the first issues we got from users could solved before delivering a prototype. The final prototype meets all requirements we planned to have and executes its task to scan a machine and show an instructional video very well. The application works on Android phones with the 4.0 version or higher and we think is a good platform to further develop. The Software Improvement Group evaluated our code and rated it at 4 out of 5 stars on their maintainability model, which means the maintainability of our total software system is above average.

Subject

NFC
scan
assistant
Scansistant
lab
machine
Android
Zend Framework 2

To reference this document use:

http://resolver.tudelft.nl/uuid:0ab38f88-0588-499a-b99e-62ee941d4fda

Part of collection

Student theses

Document type

bachelor thesis

Rights

(c) 2013 Ypeij, T.M.
Hensen, D.K.L.

Files

PDF

final_report_1.2.pdf

8.68 MB

Close viewer