A number of technologies have been trialed in recent years to provide location-based content delivery and context-aware services on the handheld platform, but the majority of audio-visual tours have used one of these solutions: 

1. Manual content navigation through a touch screen interface, virtual keypad or hardbuttons on the device 
2. Wireless LAN software-based positioning systems 
3. Infrared tags or triggers 
4. Outdoors, GPS and assisted GPS triggering

 

 

Other more experimental technologies tested include: 

1. Bluetooth tags or triggers 
2. RFID tags 
3. Ultrasonic positioning 
4. Visual recognition triggering 
5. Radio frequency triggering and delivery

 

 

 

Manual Content Navigation 

PDAs and other screen-based multimedia players can support keypads and other visual interfaces to allow visitors to access tour content manually, either by selecting an icon or typing a number on the touch-sensitive screen, or by using the device’s hard buttons. Some PDAs include programmable hard buttons or even an alphanumeric keypad; other solutions simply display graphic keypads on the screen, and users type in their selections with their fingers or a stylus. Visual interfaces include maps, room reconstructions, thumbnail icons and other menus whether text, graphic or image-based.

 

The majority of audio-visual tour projects both currently and in the past have used manual content navigation. In one of the earliest and most influential handheld projects for the ‘Points of Departure’ exhibition at the San Francisco Museum of Modern Art, visitors found the thumbnail images used as links to information on specific artworks in the tour’s interface to be clear and easy to recognize. (Smith, 2001, p. 116) Xeroc Parc’s team carefully assessed location-based technologies when designing their tour of Filioli House in Woodbridge, California, and considered IR triggering at room thresholds. They concluded, “proximity-based approaches would probably not give the user the information they wanted in the vast majority of cases.” (Aoki and Woodruff, p. 5) As a result, this project used a purely visual interface to information in each historic room; in photographic representations of each wall, the visitor could touch selected ‘hyperlinked’ objects with their stylus to trigger further information on that object or aspect of the decor. Wall views were changed using the device’s hard buttons to pan right or left around the room. Feedback from visitors confirmed, “visual selection is a viable alternative that allows visitors to quickly and easily select objects that interest them.” (Woodruff et al., 2001, p. 1)

 

At Tate Modern, a keypad interface replaced early experiments with WLAN location-based content delivery, making the tour more familiar to visitors from their previous uses of audio tours. This instant recognition proved to be essential to the successful operations of the multimedia tour of the blockbuster Frida Kahlo exhibition in 2005, where staff have only a few seconds to distribute tours and instruct visitors in their use, and is also fully accessible to Deaf users of Tate Modern’s Multimedia British Sign Language Highlights Tour. The J. Paul Getty Museum is also currently making a multimedia tour of the exhibition, Rembrandt's Late Religious Portraits, available as a locally-stored, manually-triggered tour using a thumbnail- driven interface. Visitors can download the tour to their own PDAs from the Getty website as well as pick it up at the exhibition. (http://www.getty.edu/art/exhibitions/rembrandt/download.html)

 

 

For sighted visitors, therefore, visual interfaces using the touch-screen of a handheld device can be a user-friendly and cost-effective solution for navigating through the tour. At the very least, it is a good idea to design audio-visual with tours multiple ways of accessing content in case one method fails, either because of technology or a visitor’s particular needs or proclivities. (Aoki and Woodruff, p. 7; see also the Indianapolis Museum of Art’s current PDA project) Some 

 

solutions have been developed for users with low-vision, including hard button overlays and designs for raised, rubber overlays for the device’s touch-sensitive screen; however, it may be precisely in the area of accessibility that other location-based technologies such as Bluetooth will 

justify their greater technical complexity and cost over simple manual navigation. 

  

 

WLAN Software-based positioning

 

WLAN location-based services (LBS) were first trialled by Tate Modern in their first multimedia tour pilot in 2002, and subsequently by the Royal Sonesta Hotel in Boston, and the Renwick Gallery, Smithsonian American Art Museum in 2003 and the Royal Institution in London in 2004. A wireless location-based tour was also piloted at the Posti Museum in Helsinki in 2004 and the Singapore Science Center in 2005, and wireless positioning has also been tested at the Blanton Museum at the University of Texas in Austin, as part of the GettyGuide system at the J. Paul Getty Museum in Los Angeles, and at the Museon in Den Haag in the Netherlands. Several Smithsonian Museums in Washington, DC, including the National Postal Museum, the National Museum of American History, the National Air and Space Museum’s Udvar-Hazy Center, the National Museum of Natural History, and the Castle Building (Smithsonian Information Center) planned for wireless positioning systems as part of a large-scale Smithsonian-wide pilot programme called the SIguide, and future installation at the National Portrait Gallery, Smithsonian American Art Museum, and other locations were also under discussion; however, these projects were never fully realized, in part because sufficient stability in wireless positioning was never achieved.  Also in Washington, DC, a handheld tour with WiFi positioning was implemented at the US Botanic Gardens. All of these projects in DC were designed to work with the tour content stored locally on the PDAs.

 

 

Software running over a wireless network covering the exhibition space can be used to calculate visitors’ whereabouts in the museum and deliver content automatically on the basis of their location. The principal elements of this solution are the wireless network and the software that does the location identification and application management.

 

 

 

The wireless network is composed of four main elements: 

1. The tour server, on which the server-side positioning software is installed and where the  location calculations are made. The server may be an existing part of the museum’s infrastructure, or a dedicated server can be installed to isolate the tour’s activities from the rest of the building’s network.
2. The transponders or ‘access points’, which relay radio signals between the visitors and the server. 
3. The mobile computers (PDAs or web tablets), including ‘WiFi’ antennae or wireless cards (built in or external) and the ‘client’ software for the positioning system, carried by the visitors.
4. Standard Ethernet cabling and switch(es) as required, to connect the access points to the server. 

 

Typically the access points are hardwired to the server, and in all cases the components of the wireless network require power, so ironically the only ‘wireless’ part of a wireless network is the connection between the visitor’s tour device and nearby access points. As visitors move about the museum or outdoor site, their handheld devices send signals to the access points, which are relayed to the server via the switch. By comparing the signal strength readings from access points near the visitor to a radio map of the space covered by the wireless network, the positioning software on the server can calculate each visitor’s likely location with respect to each access point.

 

 

 

The exhibition space’s radio map is created at the time that the LBS is set-up. The LBS software includes a survey tool that allows operators to record the strength of the signals received by the network’s access points from a portable device. This ‘signal strength’ or radio map is then related to an architectural map of the exhibition area, so that given combinations of signal strength readings can be correlated with specific locations in the galleries. With signals from at least three access points, the server can triangulate each visitor’s location, but even with one or two access points, the software is able to make some educated guesses about the visitor’s location on the basis of the signal strength to their portable device. Some systems, such as that used at the US Botanic Gardens, do not in fact rely on triangulation but rather signal strength reading from a single access point in each triggering space. The pilot installation at Musée d'Orsay in 2006-7 by Espro-Acoustiguide, using a proprietary wireless technology from the German company, Lesswire (akin to a Bluetooth protocol, in fact), similarly used single area beacons to update maps on visitors' handhelds to indicate which gallery the visitor was currently located in. 

 

 

The resolution or ‘granularity’ of the triangulation-based positioning platforms today is from one to three meters, depending on conditions, and most providers claim a 90-98% accuracy rate for their positioning solutions. Because WLAN solutions are based on radio signal strength, their accuracy is dependent on the conditions in the galleries and how they impact the radio signal strengths. The number of simultaneous users may cause signal strength readings to fluctuate, resulting in inaccurate positions being detected for users. Atmospheric conditions and the humidity of the local environment, which can be impacted by the number of bodies in a given room, for example, can also cause changes in signal strength readings. 

 

Most LBS software systems take upwards of 60 readings per second in order to increase the probable accuracy of the location calculations, and adjust for variations that can be predicted. They will also use intelligence from the visitor’s previous movements to exclude calculations that are wildly inaccurate or impossible – i.e. the software will assume that a visitor cannot move from one end of a large museum to another in the space of a couple of seconds, so any calculations that imply this sort of a location change are discarded as erroneous. For finer or ‘pin-point’ positioning, WLAN technologies can usually be combined with Infrared, Bluetooth or Radio Triggers, so that the software-based system handles larger, room-level racking, while the local tags or triggers provide more precise and sometimes quicker and more reliable content delivery for individual exhibits within the room.

 

The LBS software usually includes applications that can do different things with the raw X, Y, Z coordinates of a visitor’s location, such as deliver content relevant to nearby exhibits, or allow site staff to track visitor movements in real-time by watching a map of the galleries on which icons representing the visitors move.

 

 

 

Experience with WLAN software-based solutions has shown that in addition to incorrect location readings, visitors are most frustrated by the latency that seems to be inherent in location-based content delivery. Some latency actually increases the accuracy of the location-based response, as it allows the LBS system to take more readings of the visitor’s location. Additional latency can be caused by other operations over the wireless network, such as content delivery, or the handheld device’s processor speed. In practice, it takes from 2-20 seconds for the visitor to receive content when they move into a new location, whether it is requested by the visitor or automatically pushed by the server once the visitor’s new location is detected (Tate Modern Multimedia Tour Pilot 2002; Renwick Gallery pilot, Smithsonian American Art Museum, 2003-4; 

Royal Institution Science Navigator pilot 2004; GettyGuide 2004-5; US Botanic Gardens 2005). Latency and accuracy can be improved by installing more access points; both the ScienceAlive project at the Singapore Science Centre (Exploratorium, 2005, p. 11) and the Blanton Museum (Glenda Sims, comments to author 24 August, 2005) increased the number of access points on their network by 50% to speed response times, while Museon in the Netherlands had to double the number of access points on their network in order to achieve 1.5-2m granularity because of difficulties posed by their museum’s architecture (Hub Kockelkorn, email to author 19 August 2005). With their ‘GettyGuide’, the J. Paul Getty Museum in Los Angeles encountered similar challenges to wireless positioning from the site’s complex architecture.

 

 

 

Few LBS technology providers make a policy of providing museum-specific applications as an integral part of their software suite, so it is often up to the museum or another provider to adapt the LBS software to the wireless tour installation, or to use tools provided with the LBS system to 

 

‘keep the network in tune’ – i.e. optimised for location-awareness. Achieving consistent performance from WLAN positioning systems remains a challenge: as Michael Edson, Chief IT Officer of the Smithsonian American Art Museum notes, “making WiFi-based locality resolution work all-the-time (!) means spending time and money designing and refining the core wireless network, and using a wireless network management suite to tune it on an ongoing basis. The underlying wireless network has to be thought of as an integral component of the handheld solution and it has a high cost of ownership.” (email to the author 23 August 2005) 

 

 

Infrared Tags or Triggers 

 

Infrared triggering has been used for some years in museum audio tours, usually by installing the infrared trigger in ceilings or over doors to trigger content to play as visitors pass by. This sort of trigger was also used for the 2001 eDocent experiment at the American Museum of the Moving Image in Astoria, New York in 2001, the multimedia tour of the That’s Canada exhibition at La Cité des Sciences et de l’Industrie in 2004, and the 2003 tour at the Marble Museum in Carrara, Italy. A similar installation was used in the location-sensitive Telecity exhibition tour installation at the Bauhaus in Dessau in 2003. 

  

Infrared (IR) is a line-of-sight technology, requiring a clear path between the trigger and the handheld device’s infrared receiver. Proximity to natural light can degrade the quality and hence reliability of the infrared beam, but more recent generations of IR technology have proven quite robust, and can even operate out-of-doors if the user can get fairly close to the trigger.

 

 

Infrared can be used to trigger content that is either locally-stored on the tour device or delivered wirelessly through a network. Infrared can also be used to transmit content, though it is a fairly slow medium so impractical for most museum tours, which will include large size audio if not audio-visual files. The triggers can be ‘active’ or ‘passive’, in other words, they can actively trigger a mobile device whenever it passes within range of the infrared beam, or they can passively wait in an energy conservation mode for a visitor to ‘point and click’ their infrared receptor (PDA or phone, for example) at a trigger to wake it up and cause the content associated with that trigger to download. 

 

For finer, object-by-object triggering, several museums have recently experimented with small-size infrared ‘tags’ that are designed to sit next to exhibits or even be built into exhibit labels without causing too much visual distraction from the objects on display. To date, infrared tag solutions have been trialled in multimedia tours at the Experience Music Project in Seattle, the Buffalo Bill Historical Center in Cody, Wyoming, the Brooklyn Museum of Art in New York, in the UK at The Fitzwilliam Museum in Cambridge, At-Bristol in Bristol, and the National Space Centre in Leicester, as well as at The National Museum of Ethnology in Leiden, The Netherlands and the Museum of Anthropology in Vancouver, Canada.

 

 

 

The content associated with each tag can be updated and managed through the software that comes with the IR tags. Most software today allows the content associated with each tag to be managed through a browser-based application, so it can be controlled remotely and also alert the system administrator if a tag’s batteries are running low. Few systems have been in place long enough to verify the battery life of the tags, but most providers calculate from 2 months (for watch battery-operated models) to 2 years (for AA battery-operated models, depending on usage). Tags can be wired into the mains power to avoid battery problems, and/or multiple tags can be installed next to an exhibit to provide redundancy. 

 

 

IR tag solutions usually aim at providing at least one tag for each exhibit, so require more triggers than the larger, room-level granularity installations described above. In addition, it can be difficult for visitors to get a sufficiently close and uninterrupted connection to the tag if there is a large group around an exhibit. (Aoki and Woodruff, p. 1-2) These factors naturally increase the overall purchase and maintenance costs of this sort of system. 

 

 

With object-level IR triggering, the visitor must be in front of an exhibit (or rather, its tag) in order to receive content about it. With a range of 15-100cm (4-40 inches), the infrared tag’s discrete size can be a double-edged sword: objects with small tags can be difficult to locate in a gallery, and visitors have complained that this new technology does not necessarily solve the old problem of finding their way to the objects on the tour in the first place. After their early pilots, at least two museums working with IR decided to move from an object-based tag system to triggering room-level maps and thumbnail menus or a keypad through which the visitor can then manually navigate to the desired object-level content. 

 

 

GPS Triggering 

Outdoors, assisted Global Positioning System (AGPS) is the most effective positioning solution currently available. With accuracy down to one meter in optimal conditions, GPS can provide real-time tracking of visitors’ movements and respond with location-based content delivery. GPS does not, however, currently work indoors, and can be interrupted by tall buildings and heavy foliage as it is a line-of-sight technology (like IR) that requires a clear ‘view’ of the user’s GPS device by the satellites that then triangulate the user’s location.

 

At Ashton Court Manor in Bristol, England, GPS-enabled multimedia players ping users when they have entered an area that has interpretation, allowing the visitor to then play audio-visual commentaries for selected areas of the 850-acre site. A map also displays the user’s current location as a blinking red dot that moves across the map as the visitor moves through the grounds. GPS has also been trialled in a tour of Hadrian’s Wall in Carlisle, Scotland and in a number of handheld city tours, including Appleby, Lancaster University’s GUIDE project, and the m-ToGuide project tours of London, Siena and Madrid.

 

GPS is becoming more common in taxis and cars, with a range of consumer versions now available for everything from driving to hiking and sailing. The ubiquity of GPS in every day navigation has raised the bar for location-based solutions in cultural settings; often visitors ask why a museum can’t simply install GPS to help with wayfinding, and this is likely to remain the standard to which positioning systems for interiors aspire for some time to come.

 

 

Bluetooth Tags and Beacons 

 

Bluetooth tags can be used in much the same way as infrared tags, but have a larger range (up to 10m/30ft), and do not require line-of-sight. This can make Bluetooth preferable to IR as a triggering technology in situations where the visitor might not be able to find the trigger easily, either due to low vision or the organisation of the exhibit. Bluetooth is a radio-based technology, so the tags will trigger within a sphere of influence rather than in a precise, point-to-point manner like infrared; however, with a much more limited range than WLAN, the granularity of Bluetooth triggering can be refined down to 1-2 meters. Combination IR and Bluetooth tags also exist, and have been used primarily in retail and other promotional environments to ‘beam’ audio-visual content, such as ring tones, video clips and songs, to users’ own phones and PDAs. 

 

Although its shorter range and one-to-one correspondence between tags, locations and content can make Bluetooth seem a simpler and more accurate positioning solution than WLAN software-based systems, as a radio-based technology Bluetooth is subject to the same vagaries of atmospheric conditions and numbers of bodies in a given space, so signal strength from Bluetooth tags can fluctuate. When tags are located near one another, Bluetooth readers can also ‘dither’ between two or more overlapping signals.

 

 

 

Bluetooth is a relatively new triggering technology. Among the early pilots was a 2004 trial at Madame Tussauds in London. A Bluetooth 

audio tour solution was also demonstrated at the Melbourne Museum in Australia in 2003. At Madame Tussauds, Bluetooth tags were embedded in selected figures and used to trigger multimedia tour content, locally-stored on a Bluetooth-enabled PDA. Some of the portraits were located close enough to one another that visitors experienced dithering between their tags, which caused the interface to flash between the content menus for the two adjacent portraits. This is, presumably, a problem that can be resolved through judicious placement of the tags and clever design of the user interface. While perfectly responsive with a small number of simultaneous users (e.g. two), reportedly the tags sometimes failed to trigger content if a larger number of visitors gathered around the same portrait at the same time. Clearly, as with any new technology, there are platform issues to work out, and the constantly- moving goal posts of new hardware and upgrades to the PocketPC operating system have not made this any easier. However, Bluetooth is potentially a very powerful solution for location-based services for visually-impaired visitors in particular, where line-of-sight technologies are not practical and a relatively fine granularity of positioning is required. Bluetooth also has the advantage of being installed in a wide range of new consumer handheld devices, including phones and laptops as well as PDAs. This means not only that the general public is gaining everyday familiarity with the technology, but also that more visitors are likely to arrive at the museum with their own Bluetooth-enabled devices. If good solutions for either streaming or beaming content to these devices are developed in future, Bluetooth could provide an unprecedented level of access to museum interpretation.

 

 

 

RFID (Radio Frequency Identification) 

Radio Frequency Identification (RFID) is a generic term for technologies that use radio waves to identify objects automatically, but it is most commonly used to refer to systems where small, low-powered RFID chips are attached to or embedded in objects whose ‘identity’ can then be ‘read’ by an RFID receiver. An RFID chip comprises a microchip and a tiny antenna that transmits this data from the chip to a reader. The reader is activated whenever the antenna comes into range and the data can be used to trigger an event. Some museums use RFID systems to tag their objects for security and maintenance. Usually the range is no more than a few feet, and for the more common, low-powered RFID systems, only 5-25cm (2-12 inches). The short range of RFID requires the user to come very close to or touch a tag to activate the trigger, so the practical logistics of this technical constraint have to be considered in exhibition as well as tour design. Unlike IR, RFID does not require line-of-sight between the RFID chip or tag and the reader. Like WLAN, RFID systems use radio signals which can move through permeable materials, although RFID signals are usually much weaker than WLAN so can’t travel as far.

 

 

One of the earliest museums to use RFID with the public was the Tech Museum in San Jose, where children can wear RFID ‘TechTag’ wristbands that trigger exhibits and collect information for review on personal webpages. In the eXspot RFID system being trialed at the Exploratorium in San Francisco visitors carry keepsake RFID cards. (Hsi 2004 and http://exspot.exploratorium.edu) Battery-powered transceivers at the exhibits allow visitors to use their RFID cards to bookmark information on exhibits and, in some cases, trigger a camera to capture a photo of themselves or the results of their experiments at an exhibit. These photos and bookmarked information are saved in the user’s online account for later viewing. (Exploratorium, 2005, p. 23) 35% of visitors are following up on the Web or at kiosks in the museum.  (Exploratorium, 2005, p. 17) Reviewing personal photographs had a similar appeal for visitors using the IR-triggered tour in the That’s Canada exhibition at La Cité des Sciences et de l’Industrie in 2004, where visitors could take a picture of themselves in the exhibition and go to their personal webpages to see it later. Also at La Cité, barcode readers at exhibits in selected exhibitions have enabled visitors to use their admission tickets to the museum to record their route around the exhibits on personal webpages with the museum’s proprietary Visit+ system. Visitors can review their visit record online or at information kiosks in the museum. As of early 2005, 104,000 personal websites had been created for Visit+ users, receiving a total of 70,000 hits. (Topalian, 2005) The Technisches Museum in Vienna provides a similar ‘SmartCard’ service, allowing visitors to track and bookmark information from the exhibit. Like the Exploratorium, the Miami Museum of Science Planetarium is hoping to use RFID tags in its Shark Bytes Exhibit to track visitors and trigger interactive exhibits.  (http://www.prnewsnow.com/PR%20News%20Releases/Art%20And%20Entertainment/Museums/ inLogic%20Announces%20RFID%20Pilot%20Solution%20for%20Miami%20Museum%20of %20Science%20%20Planetarium%20Shark%20Bytes%20Exhibit) Granite State MetalWorks, a commercial art gallery in Littleton, N.H., has already placed RFID chips in labels next to artworks. Visitors can carry a PDA and RFID reader pen as they look around, triggering text- 

 

based information about each artwork to display on the PDA when they touch the label with their pens. The pens then Bluetooth the RFID tag information to the handheld device, which displays further information on the artwork, including price, provenance and other details from the gallery’s collection management database. It is hoped that being able to carry information in a discrete, personal format will prove more user-friendly for potential clients who might otherwise be afraid to ask questions of the gallery staff.  (http://www.rfidjournal.com/article/articleview/1540/1/9/) 

 

  

 

A kind of RFID tag is currently being used at Legoland in Denmark to help to identify and locate children lost in the park. Unusually, these tags use the same frequency as WLAN, allowing the signal to be tracked over a much larger distance, but making this in effect a WLAN technology. 

(http://www.rfidjournal.com/article/articleview/921/1/1/)

 

 

Both traditional RFID and barcode technologies as implemented to date indicate that these close- range technologies may be better suited for exhibit triggering, bookmarking and information tracking than wayfinding or other navigational assistance in the museum. Nonetheless, the action of bookmarking also seems to have a value in and of itself: as Sherry Hsi found in the Exploratorium’s eXspot project, “…the value is not in the keepsake but rather in the act of making it[.] The act of bookmarking can be useful whether or not visitors go back to it.” (Guidebook 2005,  p. 25) A forthcoming study of bookmarking from Silvia Filippini Fantoni confirms that in the act of bookmarking items of interest in the museum, visitors are often ‘voting’ for their favourite works as much as they are requesting further information to follow up from home or school. If nothing else, RFID has the undeniable power to give visitors this pleasure as well as to engage them in actively responding to their experience in the museum.

 

 

 

Wireless Audio 

Three other wireless technologies that have been trialled in audio-only tours of exhibitions represent perhaps the furthest ‘cutting edge’ of development in this sphere and merit at least a brief summary in the space remaining here.

 

 

In conjunction with the Mobile Bristol project, HP Labs in Bristol tested ultrasonic positioning as part of an in-house photographic exhibition. Providing approximately 25cm granularity, this technology requires that a network of ultrasonic sensors be suspended over the exhibition space, not unlike the configuration of IR triggers in the Telecity exhibition at the Bauhaus in Dessau. (See figure 9 above) Ultrasonic positioning was also piloted in the Mackintosh Room at the Lighthouse in Glasgow. (http://www.slis.indiana.edu/faculty/yrogers/papers/2002randell.pdf) The LISTEN project, trialled at the Kunstmuseum Bonn in 2003, used a new radio-based positioning technology in a similar overhead grid configuration to deliver audio content to visitors according to their location in the exhibition. Two different designs for the radio receivers, which were worn on visitors’ heads, were tested, illustrated below. As visitors moved through the exhibition, the audio content they heard changed according to their location.

  

 

A final highly experimental project by Simon Fraser University, “ec(h)o” at the Canadian Museum of Nature in Ottawa, uses vision- as well as position-tracking technologies to trigger ‘soundscapes’ at each exhibit. The interface to the audio content is a cube or ball whose sides 

 

represent the different kinds of audio tracks available at each exhibit. When visitors turn the object to one side or the other, a camera positioned over the exhibit recognises this gesture, triggering the system to deliver the appropriate audio wirelessly to the visitor’s headphones. 

(http://www.archimuse.com/mw2004/papers/wakkary/wakkary.html)

 

 

From the beginning, multimedia tour development has benefited enormously from research and development projects led by university groups and technology labs. The large number of these over the years is testimony not only to the enduring attraction of museums and cultural sites as test-beds for new technology, but also to the great promise that this sector holds for  future innovation.