In het project Het Leren van de Toekomst gaan Pabo-studenten aan de slag met ARLearn. Rianne Tolsma en Ellen Domke van Iselinge Hogeschool leggen uit hoe ze ARLearn zullen gebruiken.

Lees meer hierover op http://innovatie.kennisnet.nl/leren-in-dialoog-met-de-omgeving/

ARLearn is a toolkit for mobile field trips and (serious) games. The ANDROID app is already for some weeks available via Google Play. It has been successfully piloted with cultural science students in Florence and an ARLearn security simulation has been organised with employees of UNHCR.

Till now ARLearn content creation was cumbersome. Apart from manually crafting JSON encoded game files, there was no convenient way to create an ARLearn game. Today we're releasing a functional authoring tool for ARLearn. Admitted, the tool is not perfect yet but it greatly speeds up the time necessary to create a game. Furthermore, this tool supports monitoring the progress of the different users. By creating a game run, games can be played with real users.

The following video shows how to get started with the ARLearn authoring tool.

Try this yourself at http://streetlearn.appspot.com/Authoring.html and download the ARLearn android app. This release of the authoring tool features

• Four kinds of media objects: narrative objects, multiple choice questions, videos and audio objects.
• Map based (re)positioning of media items
• A powerful notification framework. Modifications to runs or games are made available real time to the android app.
• Support for downloading games and runs to your computer, so that they can be reused.

Other features that are supported by ARLearn, but have not yet made it to the authoring tool include

• Dependencies. Through dependencies media objects are linked. For instance, feedback objects can be bound to actions such as giving an answer.
• Radius triggers to make media visible only when the user is near to an object.
• Manual triggers that enable the operator to make items appear or disappear.
• Other media objects like pickup items or zones where items can be dropped.

Google App Engine (GAE), a cloud based solution for hosting servlets, features a whole range of APIs with two candidates for building a notification system:

• the Google channel API 
• Extensible Messaging and Presence Protocol (XMPP)

GAE implements a HTTP servlet container that can either fuel REST-based web services or plain HTML applications. In this internet-based communication model, the client normally initiates communication and pulls information from the server. A vast amount of mobile applications however requires push technology where the request is initiated by the server, i.e. GAE.

In ARLearn for instance, both the smartphone app and the Google StreetView application build on on push technology to notify the user when new content is available or when the (team) score has been altered.

XMPP and the channel API enable such push notifyications, but both have pros and cons. XMPP is the successor of the jabber protocol for exchanging chat messages. This protocol supports communicating plain chat messages and structured xml messages which makes the protocol a candidate for notifications. Originally, the ARLearn smartphone app implemented an XMPP client, listening for messages from the account that is associated with the GAE application. There are however many disadvantages to this approach:

• In order to communicate with the server, one needs a Google account. In the case of ARLearn, the user's Google account is used to communicate through the google talk server with the GAE XMPP account. This means that the GAE account becomes visible in your list of Google talk friends, which is a bit artificial.
• Google talk is a service hosted on talk.google.com on port 5222 or port 443. ARLearn experience has shown that a connection with talk.google.com on 5222 is blocked on many public or corporate networks.
• For StreetLearn - a JavaScript Google StreetView mashup - the situation is even worse, as the browser blocks communication to domains different from the domain hosting the JavaScript code.

The Google channel API implements long polling (aka 'Comet programming'). With this technique, the browser sends a request to the server that does live beyond the 30 seconds limit. The server holds this request and once information is available, a response is sent to the client. As the channel API was designed for web applications and targets JavaScript clients only, it overcomes many of the limitations of XMPP:

• No cross domain issues 
• No Google accounts necessary and thus no side effects like messing up your contact list. As this API features concepts like channels, client identifiers and tokens to access channels, it enables the dynamic creation of communication channels. 

There is one big disadvantage to the channel API: it only features a JavaScript client. As a result, native smartphone apps built in Java or Objective C cannot profit from this technology. For Java, an elegant short solution (4 java classes of core code) has been implemented by Tom Parker that reverse engineers this JavaScript API. This makes the channel API usable in java-based desktop applications like native android. 

Today, I published ARLearn on Google Play, formerly known as the Google Market. There are still some glitches in the tool, but we decided to give this a try and start collecting feedback. So, this is a warm invitation to send us your ideas or feature request.

For those that are new to this topic: ARLearn is a toolkit that combines field-trips, augmented reality and serious games. With the toolkit one can create “games” (e.g. a simple field-trips) and “runs”. A game is a blueprint and captures the design of your mobile activities. A game can be materialized into many runs. Within a run, a fixed set of users can act and even compete. Also, actions performed by users, response that were given, etc are all collected within a run.

The tool that we released today is fully functional. However, so far we have not released the authoring tool yet. So if you want to be on the bleeding edge, go ahead and download. Drop us a note and we will help you to get your first run installed. If you rather see yourself as an early adopter, we advise you to wait for some more days. We are now working hard now on finishing the beta version of the authoring tool for ARLearn. With this tool it will become possible to create your own field trips, play them with students and collect results.

Read more about ARLearn.

Download the application on google play.