What’s On TAP? – Strategies for keeping pace with rapid changes in mobile technology

Charles Moad and Rob Stein, Indianapolis Museum of Art, United States

http://www.tapintomuseums.org/

Abstract

In 2011, the New York Times published an article that reinforces what many museums are feeling as they scramble to keep pace with changes in technology. Entitled, “Software Progress Beats Moore’s Law”, the article cites a White House advisory report detailing the changes in hardware and software innovation in the past 15 years. The report notes that while the rate of change in hardware innovation is staggering, changes in software have vastly outpace even those impressive numbers.

Perhaps most challenging for museums today are the changes occurring in the mobile industry.  This leads to questions about how best to take advantage of mobile tools to promote access, interpretive content, and engaging experiences for the public – and at the same time, find solutions that are scalable and able to be maintained over a long period of time.

As described in (Stein 2011), the Indianapolis Museum of Art (IMA) proposed a platform called TAP that envisioned a modular software framework in addition to a formal description of mobile tour content called TourML. The IMA released an initial version of the TAP and TourML framework in 2009, which has subsequently been adopted by a number of museums. While the initial TAP software proved stable and capable of serving a variety of needs in museums, it fell short of the ability to support the vision and needs of the community as expressed in several working meetings about the specification.

In 2011, the Institute for Museum and Library Services awarded a National Leadership Grant to a collaboration of museums, led by the IMA, to create and extend upon the TAP & TourML open-source platform to create a suite of interchangeable tools that can support the creation of reusable mobile experiences. This paper will discuss the goals of the project, the current state of the TourML specification (including its relationship to other open-source mobile projects), and will outline the software development roadmap for the project. Authors will describe the various components of the TAP system and how modules from the toolkit can be used individually or in combination to build a variety of museum mobile experiences.

Rather than being a stand-alone product, the paper will document the ways that TAP and TourML seek to integrate with existing museum authoring tools and commercial mobile systems creating a method for interchange and migration of mobile content between these systems.  This same interchange method can help museums preserve and reuse the content developed for mobile tours and can potentially lead to the ability for museums to share and trade content elements between distinct mobile experiences.

Keywords: mobile content, sustainability, open source, standards

1.     Introduction to TAP and TourML

In 2011, the New York Times published an article that reinforces what many museums are feeling as they scramble to keep pace with changes in technology. Entitled, “Software Progress Beats Moore’s Law”, the article cites a White House advisory report detailing the changes in hardware and software innovation in the past 15 years. The report notes that while the rate of change in hardware innovation is staggering, changes in software have vastly outpace even those impressive numbers.

Perhaps most challenging for museums today are the changes occurring in the mobile industry. In 2011, Pew reports that 83% of U.S. adults now own mobile phones, and fully 44% of those surveyed used their phones to access the Internet (Smith, 2011).  In addition, the availability of easy-to-use apps on mobile devices has enabled information access and content sharing with a real-time component that transforms the way we use the web. Adoption of mobile apps has surged, doubling in just a two-year span between 2009 and 2011 (Purcell, 2011).

With 4 billion mobile phone subscribers worldwide, it is clear that mobile devices and content will be an important means of access for museum visitors today and in the future. More recent anecdotal evidence suggests that these trends have accelerated rather than subsided, and that an increasing number of museums are contemplating how they might deliver content via mobile devices. The Museums and Mobile Survey 2011 indicates that over half of large museums (annual attendance of more than 50,000) already have mobile experiences, and almost 70% of all museums say that their institution will “definitely” have in-house mobile content development within the next five years (Tallon, 2011).

This leads to questions about how best to take advantage of mobile tools to promote access, interpretive content, and engaging experiences for the public – and at the same time, find solutions that are scalable and able to be maintained over a long period of time.

As described in (Stein, 2011), the Indianapolis Museum of Art (IMA) proposed a platform called TAP that envisioned a modular software framework in addition to a formal description of mobile tour content called TourML. The IMA released an initial version of the TAP and TourML framework in 2009, which has subsequently been adopted by a number of museums. While the initial TAP software proved stable and capable of serving a variety of needs in museums, it fell short of the ability to support the vision and needs of the community as expressed in several working meetings about the specification.

In 2011, the Institute for Museum and Library Services awarded a National Leadership Grant to a collaboration of museums, led by the IMA, to create and extend upon the TAP & TourML open-source platform to create a suite of interchangeable tools that can support the creation of reusable mobile experiences.

2.     Goals of the Project

While the overarching work of the project is diverse and wide ranging, the formal goals set forth focus in three main areas:

1. Create open-source tools that can be used in full or in pieces to support the creation of mobile experiences in museums;
2. Build community support and adoption of the TourML metadata specification as a means of representing the content of mobile museum experiences;
3. Successfully use these tools and methods to launch and document mobile experiences in each of the partner museums.

It is the vision of the project that the software created will be designed as a suite of tools that can form the component parts of larger and more custom mobile tour experiences. While it will be possible to create a complete mobile experience using the open-source tools provided by the project, we anticipate that most users of the toolkit will benefit from the software provided by integrating it into apps and websites that are unique to their museum.  Indeed, we anticipate that many commercial vendors will benefit from having access to free software tools that help ease the path of integrating their commercial systems with content authoring systems, or front-end user interfaces. This flexible approach should ensure the optimum utility of this software for use by museums.

The TourML specification promises to be an important long-term contribution to the ability for museums to preserve and migrate their mobile content as technology changes. Many museums have experienced losing the ability to extract and modify content embedded in kiosk or other proprietary software tools. TourML seeks to help with this problem by describing the structure and connections present in mobile content using a platform neutral XML schema that can be parsed and used by a variety of current and future software tools. Furthermore, the effort seeks to benchmark the common needs many different museums have for creating this content and building consensus regarding how that content should be defined and described. The details of the TourML specification will drive the software features required in content authoring and user interface tools. It is the hope of the project that a variety of software vendors will see a benefit in integrating support for the TourML specification into their products.

Lastly, the project plan includes a deep collaboration with a number of participating museums to help in designing both the software products, and in building an appropriate specification for TourML. As an end product of the grant, each participating museum will use TourML and optionally components of TAP to host a mobile experience for their museum. Each will also author a whitepaper that reviews the process and experience in using those tools given their perspective as a museum in the community. In addition to serving as a valuable evaluation of the projects goals, we also hope that these whitepapers will serve as useful case studies describing how the tools of the project can be deployed in museums in a number of discrete ways.

3.     Current State of TourML

The project has hosted several workshops with media and technology staff from the cultural sector as well as commercial vendors of mobile software in order to understand how TourML can best represent the mobile interpretive content that is being generated by museum. Based on this feedback from the community, the TourML specification has rapidly stabilized. A 1.0 version of TourML is planned to be ratified before the end of 2012. Today, TourML has three primary components for representing mobile tours and their content, and two secondary components for adding metadata to tours and for the combination of tours in order to create sets of tours.

Figure 1: A sample diagram of Tour demonstrating the relationship between Assets, Stops, and Connections.

Assets

Any piece of content that is vital to a tour is considered an asset. This includes traditional media such as video, audio, and images. Also included is the narrative text that appears in the tour to the visitor. Anything that is crucial to the meaning of the tour content is considered an asset. This principal has served as a guideline for how to handle less obvious components of a tour.

Geolocation information is extremely prevalent in today’s mobile tours. So much so, that the initial approach for TourML was to include geolocation data directly on tour stops. As the project team explored the usage of location information they discovered that it is being used in many ways. It can represent where a tour should be experienced. It can direct users how to physically navigate a set of content. It can also be used to define boundaries in order to notify users of content when they cross a boundary threshold, also known as geo-fencing. Realizing that one cannot assume how geolocation information might be used, put it in the same class as other content assets in TourML.

There are also many examples of content that are not crucial to the meaning of the tour. Frequently this information falls into a class of machine data that is useful for driving the user-experience or navigating tour content. The removal of this content would not change the meaning of the tour, but rather only how that tour is presented. Key codes are one example by which users type into a keypad in order to access a stop. Switching around key codes on stops would not change the meaning of the tour content, whereas changing geolocation points on stops would greatly change their meaning. For this type of information, TourML includes a general-purpose element (PropertySet) that can be used to represent this additional content on each asset, stop, and tour.

<tourml:Asset tourml:id="img-1">
    <tourml:Source tourml:format="image/png"
     tourml:lastModified="2011-09-29T12:01:32"
     tourml:uri="file:///images/ankh-ct.png" />
</tourml:Asset>

Figure 2: An example Asset XML element from a TourML instance.

Stops

All tour content can be organized through the concept of stops, which the project team has defined as, a set of assets that are meant to be experienced together. This fairly open definition allows users of TourML to easily express the many types of traditional tour stops that we are familiar with today and those that we cannot predict in the future. Stops contain a set of references to assets. Each reference describes how that particular asset is meant to be used within the stop. This semantic description of assets is one of the most important roles of stops in TourML.

The ways stops are displayed can be specified by indicating a “view” which describes how a stop should be presented in an end-user application. It is up to each application to decide exactly how to present the assets within a stop for a particular view. As technology changes, so could the details of how an asset should be presented. This concept comes from a software architecture paradigm called Model-View-Controller, where applications and their content can be discretely represented in three parts. Applications that drive mobile tours must decide how users view content and control how they interact with it. TourML has been conceived as a common model that allows the cultural community to easily share their content with each other and move that content between application platforms as technology changes.

<tourml:Stop tourml:id="stop-1" tourml:view="ImageStop">
    <tourml:Title xml:lang="en">
        Ankhaman's remains
    </tourml:Title>
    <tourml:AssetRef tourml:id="img-1" tourml:usage="primary" />
    <tourml:PropertySet>
        <tourml:Property tourml:name="code">100</tourml:Property>
    </tourml:PropertySet>
</tourml:Stop>

Figure 3: An example Stop XML element from a TourML instance.

Connections

The third primary component of TourML describes how stops are related to each other. Connections describe one-way relationships between stops of a tour. Any tour can be expressed as a graph whose nodes represent stops. Connections are the edges in that graph that describe how to get from one stop to another. Connections are quite simple and may only contain four pieces of information. A reference to the source stop and the destination stop describe the direction of the relationship. A connection can optionally contain a usage, which describes the meaning of the connection. This is similar to how stops reference the usage or meaning of assets. Lastly, the connection may declare a priority. This allows for the ordering of many connections from the same source stop to different destination stops.

<tourml:Connection tourml:srcId="stop-1"
 tourml:destId="stop-3" tourml:priority="0" />

Figure 4: An example Connection XML element from a TourML instance.

Tour Metadata

All tours contain basic information for describing the set of the content they encompass. This metadata includes traditional information such as title, description, author, and when the tour was published. Additionally, users can describe assets that are to be used across the entire tour. Examples of this might include an application icon, splash screen, or sound effects that are used inside the application. Also, users may reference a single stop that points to the tour stop that users automatically visit first in the tour. This root stop reference allows tours that contain a linear navigation model to set the starting point.

Tour Sets

TourSets allow users to combine many tours together or break apart an extremely large tour into smaller chunks. TourSets address the same problem found in other XML specifications for representing large amounts of content. Examples include the Sitemap (http://www.sitemaps.org/) specification and OAI-PMH (http://www.openarchives.org/pmh/).

Each TourML document must contain either a Tour or TourSet as its top-level element. TourSets contain nothing more than a list of links to other TourML documents. Those documents must also contain a TourSet or a Tour. This feature of TourML can support the representation of tour content at a virtually unlimited scale. Every tour must contain an identifier. Tour elements that contain the same identifier describe the same collection of content that compose a tour. This also allows users to separate their Assets, Stops, and Connection among many TourML instances. This feature is critical for representing the vast object collections found in many museums that can scale to the millions.

4.     Software Development Components and Roadmap

TAP uses of a wide variety of open source software tools in combination to provide a simple suite of tools to interact with TourML documents. Users of TAP can pick and choose components that suit their individual needs as it has been designed as a modular toolkit for the authoring, consumption, and validation of tour content as described by the TourML specification.

TAP Authoring Tools

A set of tools for the creation of tour content has been developed on top of the content management system (CMS), Drupal 7. Drupal was chosen as a platform due to its prevalence in the cultural sector and familiarity by the project team. Additionally, Drupal excels at the creation of custom content types for representing individual user’s information without the need for writing any code. Due to the diversity of their collections and programs, cultural institutions frequently express unique needs when presenting their interpretive content online or on mobile devices.

One powerful feature of Drupal is that web administrators can customize the display of their custom content types through a web interface provided by the CMS without writing code. Leveraging this feature, TAP uses the robust rendering features of Drupal to encode valid TourML XML output instead. This approach allows web administrators to map the fields of their unique content types to stop elements in a TourML document. TAP provides valid XML representations for all the possible types of fields that are included with Drupal, making it easy to export existing content contained in a Drupal website to the TourML specification.

Additional support has been provided for mapping additional fields from other third-party modules. For example, the Geofield (http://drupal.org/project/geofield) module provides a powerful interface for adding geolocation information to content in Drupal. Users can author point, polyline, and polygon geoinformation by using simple point and click maps provided by that module. The TAP authoring tools allow web administrators to map these location fields to TourML assets, which can represent the location information in one of several popular formats. This Geofield specific feature of TAP serves as an example of how to map other third-party fields to TourML.

TAP provides several ways to export TourML and the tour content from the CMS. Users can export a complete TourML XML document directly from a tour. Support is also being developed for exporting bundles of content, which include the TourML XML document and all the asset files that are referenced in the tour. These bundles will be based on other existing specification for packaging content. Two possible examples include the Open Container Format (http://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm) used for EPUB documents, and Apple Packages (http://en.wikipedia.org/wiki/Application_bundle), also known as bundles. By leveraging existing packaging formats, TourML bundles can better interoperate with existing platforms. One prominent example of this is the localization support of Apple Packages, which allows for transparently loading assets in a preferred language in Mac OSX and iOS application.

Localization support is provided as part of the core release of Drupal 7. The TAP authoring tools leverage this existing functionality in order to provide translated fields of information on stops. Web administrators can choose which fields of information may be translated on a stop and which are not. The CMS manages the translation and warns if a translation might be out of date. For example, you might provide multiple translations for the title of your stop, but have a single image that is reused across all languages.

TAP End-user Applications

As part of the TAP toolkit, two example applications that are driven solely from content described by the TourML specification are going to be provided. These applications are being provided for several reasons. First, the applications serve as a demonstration of best-practice use of the TourML specification. Second, the applications themselves are being developed in a modular way so that the code for parsing and working with TourML documents can be reused in other applications developed on the same popular platforms. Finally, it is a goal of the project to provide a complete front-to-back and open source solution for delivering mobile tours. These tools can be used by institutions that cannot otherwise afford commercial mobile solutions and/or would like to rapidly prototype a mobile tour for audience evaluation before committing to a final product.

The first iteration of TAP included an iPod Touch focused iOS application with keypad style navigation and support for videos, images, audio, web pages, and user polls. This application has been adopted by a handful of museums for use on devices provided by the museum and on visitors’ personal devices on the Apple App Store. As part of the project plan, this application is going to be updated to support the TourML specification improvements based on the community feedback. The only part of the application that must be updated is the code which is responsible for parsing TourML. The rest of the application will remain the same yet be enhanced to support new higher resolution devices such as the iPhone and iPad. Additional plans for the iOS application include support for multiple tours, new styles of navigation, and additional stop types.

In addition to the iOS application, a new web-based application is being developed in order to demonstrate the flexibility of TourML and to support a wide variety of devices. This application is being built on top of the jQuery Mobile (http://jquerymobile.com/) web framework. jQuery Mobile is a well-supported HTML5 based mobile platform that has extremely broad support. It includes many rich features such as user interface components that gracefully degrade on older platforms and the ability to quickly roll a custom theme.

TAP Validation Suite

The project team feels that it is extremely important to have a tool available for validating TourML documents. Such a tool should not only validate the syntax of a TourML document, but should also inspect the semantics of the tour itself in order to evaluate whether the information provided is complete, warning users of possible limitations they might face when migrating their tour content to new systems. This tool will be built separately from other aspects of TAP in order to ensure that TourML documents coming from the TAP authoring tools face the same scrutiny as those coming from other systems. The validation suite will be a simple web page with an upload form for allowing non-technical users to evaluate TourML documents they export from third-party systems, ensuring their content is being represented adequately in order to ensure its sustainability.

5.     Integrating TAP with Museum Information Systems and Commercial Systems

The TAP toolkit and TourML have been designed in such a way that they can easily interoperate with other museum software tools and commercial mobile tour solutions.

Many museums are already moving towards integrating collections with their website CMS systems. By building the TAP authoring tools as a simple extension of the core functions of Drupal 7, it has the added benefit that it can be installed on any existing Drupal 7 based website. This would allow web administrators to map their existing content types to TourML and then have the ability to easily export that content to TourML. The project also seeks to integrate the TAP authoring tools with at least one commercial digital asset management system and collection management system. These integrations would add the ability to reference digital assets and object metadata assets from stops in the same way an uploaded video or image might be.

The second year of the IMLS TAP & TourML grant involves the deployment of mobile tours by all participating partner institutions. The project plans to exercise a wide variety of deployment scenarios, including some that may not involve TAP at all but instead use TourML to drive commercial applications from core museum systems. Other deployments will use the TAP authoring tools to create and manage tour content but deploy to third-party applications. Finally, the project team will work with commercial vendors in order to add support for TourML to their content management systems and to aid in migration of content between TAP and their systems.

6.     Future Work

It is the hope of the project that as a result of the efforts described here, it will be much easier for museums to create and effectively maintain mobile content about their collections and programs. Success will be defined ultimately by whether or not a community of dedicated users begins to grow and contribute to the continued development and enhancement of the system, and by whether or not the vendor community integrates support for the system and tools into their products moving forward.

The TourML specification as a means to describe tour content paves a path of sustainability through the ever-changing technology landscape. The ability to migrate museum content to new technology platforms as they emerge and become more prevalent is an important missing piece in the discussion about mobile content. Given the rate of change that we are experiencing with mobile technology, it is increasingly likely that without such a path it will become more and more difficult to reuse the stories told by museums. A common platform for building and talking about the mobile content we create would be extremely beneficial to the museum community. Such a common reference point would enable a variety of opportunities for sharing content and exchanging media about collections.

7.     References

Lohr, Steve. "Software Progress Beats Moore's Law - NYTimes.com." Technology - Bits Blog - NYTimes.com. New York Times, 7 Mar. 2011. Web. 26 Sept. 2011. http://bits.blogs.nytimes.com/2011/03/07/software-progress-beats-moores-law/

Purcell, Kristen. 2011. "Half of Adult Cell Phone Owners Have Apps on Their Phones." Pew Research Center's Internet & American Life Project. Accessed January 9, 2012. http://pewinternet.org/Reports/2011/Apps-update.aspx

Smith, Aaron. 2011. “Americans and Their Cell Phones.” Pew Research Center's Internet & American Life Project. Accessed December 12. http://pewinternet.org/Reports/2011/Cell-Phones.aspx

Stein, R., and N. Proctor, TourML: An Emerging Specification for Museum Mobile Experiences. In J. Trant and D. Bearman (eds). Museums and the Web 2011: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2011. Consulted September 26, 2011. http://conference.archimuse.com/mw2011/papers/tourml_an_emerging_specification_for_museum_mobile

Tallon, L., 2011. Museums and Mobile Survey 2011 http://www.museums-mobile.org/survey/