The Open Wonderland COMSLIVE project

From Birmingham City University, Nigel Wynne describes the COMSLIVE project in answer to questions for snapshot #9 (out shortly). This is a collaborative project between two universities and an NHS Trust. It’s a good reminder that Second Life is not the only virtual world with practical applications, and that funding bodies such as the JISC are sensibly not putting all their eggs into one (virtual) basket.

* * * * *

What are you doing in virtual worlds? Teaching, learning, research, publicity, and/or anything else?

Our virtual worlds project, Communication Skills Learning in Immersive Virtual Environments (COMSLIVE) is funded by the JISC Learning and Teaching Innovation Grant. The project seeks to explore the feasibility of using the Open Wonderland (OWL) virtual world platform to provide a creative and innovative solution to more effective communication and team working skills development amongst health care learners. Deficits in communication are known to be a major contributor to much of the unintended death and illness that occurs within the UK and US health care sectors each year. This provides educators with a strong social imperative to explore more effective solutions towards developing team working and communication skills.

Some readers will know that Open Wonderland (OWL) is a 100% Java, open source virtual worlds toolkit, that is free to download and incurs no license fees. The Wonderland platform was originally developed by Sun Microsystems. During our project Sun was acquired by Oracle and Oracle withdrew direct support from the Wonderland initiative. This obviously was a concern to us as for a while the future of Wonderland seemed unclear. However, Project Wonderland became Open Wonderland, the source code forked and completely transitioned to the open source community and is now steered by the Open Wonderland Foundation.

Interestingly, since this transition the forums and community have become more active, and the release of updates and enhancements has never been more frequent. OWL, therefore, is going from strength to strength.

Prior to starting our JISC project we scoped a range of virtual world platforms (this scoping report can be found on the project website) and concluded that in terms of having the greatest transformative potential to meet our needs to enhance communication skills development, Wonderland was the most promising technology that we had assessed. Some of the reasons behind our decision included:

• Open art pathway.
• Sophisticated communications tools including spatialised sound and telephony integration.
• Highly extensible architecture.
• LDAP integration and ability to deploy and host behind firewall.
• Community support.
• Drag and drop PNG, JPEG, and 3D model functions.
• Potential to scale to support large cohort learning.

COMSLIVE aims and outcomes were informed by some of the questions we felt teachers, managers and ICT staff would ask when considering the strategic rather than ad hoc deployment of virtual world learning across curricula. These included:

• How easy is it to create or use existing artwork for Wonderland?
• What levels of interactivity can be supported in a Wonderland world?
• What considerations, both logistical and pedagogic, need to be made when integrating Wonderland learning in module programmes?
• To what extent will Wonderland scale to support potentially high numbers of concurrent learners?
• What impact would our COMSLIVE scenario have on learner experience?
• What impact would COMSLIVE learning have on learner real world behaviour?
• What are the server, network and client side resource implications of running our COMSLIVE scenarios?
• How feasible would it be to deploy Wonderland within the NHS to support work based learning?

Going well? Not? Want to say why?

Yes – in short the project is going very well. (A video walkthrough of an early iteration of our world is online). We’re approaching the tail end of the project now which officially runs from Nov 1st 2009 to Oct 31st 2010. At a strategic level it has informed the procurement of new on-campus student learning, training and research PC’s so that we now have a suite of approx 120 computers that are virtual worlds capable.

We’ve been impressed with the ease of art creation, and the drag and drop functions have made aspects of scenario design almost effortless. We have run our first iteration of an RCT trial that explores impact on real world behaviour. Learner experience was very positive, and intervention group behaviour seems on initial analysis to have been positively affected by the COMSLIVE experience. We are conducting further analysis and 1-2 more iterations of the RCT between now and December. Evaluation reports will be published on the website over the coming weeks.

Some interesting and unanticipated findings have emerged from our COMSLIVE work. The first is the impact on the student – teacher relationship when learning in-world. Students and teachers both felt like they were operating on a much more level playing field when in-world, with students being much more questioning and responsible for their learning. We also found that virtual world learning seems to have a powerful impact on “friendship forming”. i.e. all of the students forming our intervention group were randomly allocated to their small group teams and didn’t know their fellow team mates. All stated that they had formed friendships and valued the opportunity to work intensively with each other and thereby get to know each other better. We now think that virtual world learning may have an important role to play in first semester teaching if only from a social networking perspective. We are considering deploying Wonderland to further enhance our strategies for minimised attrition during what is a challenging time for many students.

Money is tight. The ‘golden age’ of education money may be ending. How are you getting funded? How do you think your virtual world activities will be funded in the future?

We were awarded 75K from JISC and made a significant match contribution from Faculty and central university pots. We already have in place a learning technology team based within the Online Simulation and Immersive Education Research Group, within the Faculty, and so have native capacity to deploy and develop further. With no licensing fees to worry about we can concentrate on design, integration and research into impact. However in order to promote interactivity within Wonderland we are keen to secure JAVA skills and are therefore looking to develop collaborations with other organisations. We will also be funding a PhD studentship that will involve programming developments within Wonderland, so if anyone out there is interested please let me know.

Long distance travel is increasingly precarious. Ash, strikes and airlines going under ground flights. Travel is expensive (even in the UK with extortionate train fares) and takes up a lot of time. Virtual worlds could, possibly, be used instead of many workshops, conferences, meetings et al. Your thoughts on this? And how do virtual worlds stack up against other event-replacing media such as Elluminate and Skype?

I guess it’s the purpose of the travel that informs judgments about whether certain technologies may afford cost effective alternatives. If your aim is to communicate, collaborate, develop and share with colleagues virtual worlds are now offering powerful solutions that can make this work more cost effective. If you are largely looking at content transmission albeit with voice or video supported dialogue then Skype, Illuminate, Wimba etc. may be the way to go.

For my mind virtual worlds are a very useful simulation tool. Our interest in Wonderland was its potential to become one of a number of simulation modalities that we are integrating across our curricula; maybe 2D virtual patients, 3D virtual worlds or games platforms etc. actor focussed or computerised mannequin focussed simulations. The point for us is that each of these modalities has its own affordances and limitations. Where some provide higher fidelity, they may be geographically fixed, resource intensive and difficult to scale. Where others have lower fidelity, we may get cheaper running costs, yet more distribution capacity.

The challenge for us is to engage less in head to head evaluations but look more closely at combining and aligning modalities so that we optimise their respective affordances and accommodate for their respective limitations. We call this approach Constructively Aligned Multiple Modality Simulation. There are some presentations available (under Dissemination) that describe this approach and a paper will be published shortly.

Are you sticking with this particular virtual world?

For the foreseeable future we are sticking with Open Wonderland. Its extensibility makes it a very exciting platform to work with. For example, the VNC viewer module is a recent development that means we can display and control a desktop and any apps displayed within it, in–world. We’re looking forward to having some fun with this in the near future.

Problems with universities blocking access to virtual worlds. Is anyone still having that, or are we over it now?

Firewall issues were a consideration for us when we chose Wonderland. We wanted to be able to deploy the technology behind our firewalls and ensure that students could log in as themselves. It was also important for us in terms of communication skills training that students were not anonymous, but that they could also rehearse these skills in secure environments. With OWL we have LDAP integration which helps with these issues.

Handling large numbers of students in virtual worlds simultaneously i.e. more than 30. Do you have experience of this? How did it go?

We have a server architecture that includes an application server, and four blade servers (nodes). Three of these nodes each support a world, and within each world we can create multiple world instances. The fourth server supports Asterix, a soft phone server that provides telephony support. In a recent test we launched 60 avatars in-world and then distributed these to each of the sub world instances. Performance in each sub world was excellent. We aim to continue with these tests by doubling the numbers of clients and trying different deployment configurations. i.e. more sub worlds in one server node, using more than one node, each with sub worlds. There is real potential to now create a range of concurrent inter-professional virtual world scenarios and simulations that interconnect and support high numbers of concurrent users. Imagine, therefore, a virtual community populated by learners representing a range of agencies.

What do you think of the new Second Life viewer, both the UI/usability changes and the new functionality it enables (e.g. media on a prim)?

We took a look at this when testing out new PC specs. I though the new UI was very impressive!

Do you have a view on the new Second Life Terms of Service conditions and ownership rights which are creating a bit of a hoo-hah in some quarters? Do you think it will affect you? Does it matter in the grand scheme of things?

Not really had a chance to get my head round this. We were concerned that spending time developing resources for Second Life may be a wasted effort if it went belly up. I understand that we can re-use 3D resources created in Second Life in Open Sim, but at the time Open Sim had no native VoiP support; the platform wasn’t going to be a solution for the type of skills training we were focussing on.

With the open art pathway in Wonderland we can now create 3D assets using standard 3D authoring software packages such as blender, sketchup, Max and Maya. As we have three work streams, VCC, Unity, and COMSLIVE, all of which do or will use 3D media, we can support a fairly efficient workflow in which we can develop almost simultaneously for all work streams.

If anyone wants to know more about COMSLIVE please feel free to drop me a line at: Nigel.wynne@bcu.ac.uk