Last modified on 8 June 2009, at 02:42

Software Engineering with an Agile Development Framework/Whole process/MOAS

Map symbol attraction 02.png Case study MOAS - Virtually There

Group: Justin McCormack, Chris Rosescu, Peter Garrett and David Youngman for Otago Museum



What are the characteristics of a successful capstone project?

A combination of a good team and strong methodology combined with creativity and technical skills produced a solution that exceeded all expectations.

A strong relationship with a client is critical to the success of a project. This was the first project we undertook with the Museum, but has led to many more.

The Otago Museum was established in 1868 and is the key culture and heritage organisation for the Otago region. With a collection of over 1.7 million items, the Otago Museum is one of the four largest museums in New Zealand. With a long term duty of care responsibility, the vast majority of these items are in storage but this does not sit freely with a mission statement including “to inspire and enrich our community, and enhance understanding...”. A solution for this contradiction was the challenge set for this project group in 2002.
Moas-montage.jpg
Moas kiwi.jpg

The Museum had previous experience with 3D capture of artifacts but it was prohibitively expensive and performed on a individual piece basis. The solution is essentially a 3D capture and display system. Although the turntable was the most visible part, the true strength of the project was the fact that an integrated system was produced. To reduce the effort and time required for each object to a matter of seconds, required an integrated yet modular system. The solution is not “rocket science”, the only real novel solution is an optical switch to control the camera, yet, as Loudon (2001) argued, sometimes it is the robust and creative application of old ideas in a new way that provides innovation.

Moas timeline.png
With this project we pioneered the Agile Development Approach. In the case of the hardware system this was seen is several stages. The initial testing was on a converted record player, the test items chosen to reflect difficult artefacts (a shiny pink pig and a rock with almost indiscernible striations).
Moas2.jpg

A virtual prototype was used to establish design parameters for lighting and focal arrangements. The first production prototype was used in the client’s environment before the production version of the turntable was manufactured to industrial specifications. The timeline of the project demonstrates the importance of combining agility with a structured methodology and clearly identified deliverables and milestones.

Prototyping was also used in the development of the software applications, to both test development approaches and as a communication tool for the clients. These were, in the main, revolutionary rather than evolutionary prototypes although the strict coding standards applied by the group meant that even early prototypes were relatively robust.
Moas-prototype.jpg
Moas inside.jpg

The museum project (described in more detail in Garrett et al. 2003) aimed to provide a solution for the Otago Museum. This was a complex project that aimed to resolve the differences between two high level goals of the museum – the long term storage of artefacts that conflicts with the public education imperative. The group developed an automated system that captures 3D objects and displays them on the web.

The group performed extensive prototyping and testing as part of a carefully managed development plan. A functional prototype involving hardware and software, along with a great deal of ‘sellotape’ was developed as part of analysis. Despite considerable frailties and faked components this was tested at the Museum in order to demonstrate the worth and feasibility of a system as described by the functional requirements. Following a revolutionary pattern the group developed several hardware and software prototypes, each of which they extensively tested at the museum. Doing this enabled the group to perfect the process before automating each step. The group’s reflective review describes the excitement of capturing a first museum object (a moa bone), a tiny object (a weta) and a huge object (a stuffed tiger). Three months before the end of the project the group was able to begin the construction of a production system. This they did to perfection, both the software and hardware systems were extremely well engineered: elegant and commented code and bullet-proof and redundant hardware systems. This system was installed and in production in the museum a month before hand in date.


The finished system has several software components to capture, manipulate, store and present both the 3D artefacts and a virtual tour of the Museum. This was achieved by the development of an intranet style application that integrates MOAS components (capture, management, and database) and their output (36 images). The interface that fits with the existing Otago Museum website, logos and other publications and is designed for a broad audience. The database supports an unlimited amount of object categories with unlimited levels of hierarchy and is presented in a way that allows users to easily navigate large numbers of objects and to interact and manipulate the digital objects. Using SQL and Active Server Pages the system implements recursive queries to manage large taxonomy of object categories and generates dynamic and complex search queries based on many different object attributes. An object viewer was developed and implemented in JavaScript with additional Visual Basic Code to watermark the images. The object viewer includes a pre-loader, zoom controls, rotation controls, pan controls and toggle rotation controls.

This project was an outstanding success, and continues to be so. Fincher et al. (2001) state the importance of finding good clients, providing a real service to the clients but of the client’s having realistic expectations. In this project we could not have wished for better clients, and, at the time of delivery the client was very pleased, the project “exceeded expectations”. All four students graduated with the “Award for Excellence the B.InfoTech”.

Moas-group.jpg

The project also managed to achieve the additional benefit of creating significant exposure for both institutions.

Perhaps the most important factor of all is the group itself. They operated in a cohesive unit while assigning individual responsibilities and showing much individual flair.


References

Fincher, S., Petre, M, and Clark, M. Eds. (2001). Computer Science Project Work: Principles and Pragmatics. London, Springer. 267p

Loudon, A. (2001). Webs of Innovation. London, Pearson Education.245p