Gallery One, Technology and Data Infrastructure

A Single Version of the Data: Collection Wall, ArtLens, and Website

The dedicated Piction CMS built to support the Collection Wall and the ArtLens app creates 23 specially-sized derivatives (11 pairs for iPad use, and one for wall display) from the master images stored in the main Piction DAM used for internal research. In the first year of operation, IMTS had to add a workflow for reviewing the horizontally-cropped derivatives used to advertise media tours; autocropping sometimes resulted in a Ken Burns-style zoom straight to Rubenesque cleavage or Apollo’s groin.

In 2013-2014, IMTS made some fundamental changes to the Piction DAM as well, ‘flattening’ the artwork-photography collection so that each image is represented by a separate record, to support image-level cataloging and manipulation and better access to alternate views. With this project, the entire artwork data flow—from CCMS through three Piction DAMs, and ultimately to the Collection Wall, ArtLens, and Collection Online on the museum website—was re-coded for efficiency and thoroughly tested for different scenarios of metadata and image.

Aligning the museum’s backend systems

Making sure the content on the both the Collection Wall and ArtLens was dynamic and maintainable was especially important to the museum.

All information is pulled directly from in-house digital asset management systems, so any newly-acquired artwork (or an artwork that has gone off view) is immediately incorporated into both. The development of Gallery One and the ArtLens app forced the alignment of two primary, integrated system backbones at the museum. The collection-information backbone encompasses all cataloging and interpretation for the artworks in our collections, studio photography of the artworks, and all the bibliographic and archival resources available to support ongoing scholarly research. The visitor experience backbone captures ‘user’ moments (i.e., transactions) that document a person’s relationship with the museum—from parking-garage entrances and special exhibition entrances, to attendance at MIX cocktail nights, and children’s art-class registrations, to donor-circle renewals and volunteer shift check-ins. In the first year of operation, both systems integration and data integrity have been hardened to provide a reliable, efficient source platform to support multi-channel access.

The Collection Wall’s cross-collection display themes and coverflows of related artworks are dynamically generated based on metadata. Some of this metadata flows directly from the museum’s collection cataloging and management system (CCMS), while some is curated and added to use-specific content management systems. During year one, the metadata structure and content were expanded to include short names for the long-winded curatorial collection titles, sortable artwork dates, standardized statements for materials and techniques; and creator and donor statements were scrubbed for consistency. To support the creative use and application of the museum’s abundant information about its permanent collection and art in general, IMTS had to build a solid backbone for collection information. The effort is ongoing an all fronts, but CMA took a significant step regarding the digital asset management component of that backbone this past year.

Content is ‘Everything’

Visitors’ use of ArtLens in the galleries set in motion a concerted effort to expand ‘label’ information for artwork on view. Curators and Interpretation staff wrote narratives for artworks that had none, filling in long-known but previously unimportant gaps, and quickly realized that context was everything: What a visitor could glean reading a wall label in the galleries was wildly different from what he would understand looking at the same artwork’s page on the website. With ArtLens, visitors could be standing in front of a sculpture, or sitting in a Starbuck’s browsing the collection on an iPhone.

New fields were added to multiple systems in the collection-information backbone, to support mobile-device specific narratives (called App Text), and to support new features introduced with smartphone app development:

  • markers used to identify curatorial ‘top 10 picks’ throughout
  • artist names without dates or nationality for efficient

AV Integration

Throughout the research and development process, the Gallery One project team quickly recognized that early coordination of the AV hardware and computer systems was essential to keeping the design on budget and schedule.

CMA hired an AV integrator. Early participation of the AV integrator in the design phase allowed coordination with the architects and electrical and mechanical engineers to optimize the space build-out and installation, including:

  • Power requirements and locations and types of circuits.
  • Conduit specifications for AV systems. The AV conduit was installed with the electrical to reduce costs and duplication.
  • Heat loads were calculated early in the project to size heating, ventilation, and air conditioning for electronics that generate heat
  • Coordination with the architect was critical to minimize the impact on the space of devices that generate noise.
  • Early coordination located the floor penetrations and ceiling devices to minimize impact on the space.
  • Early coordination on the mounting locations—especially for the MicroTile wall—was critical because of the tolerances of the wall being less than 1 millimeter of variation across the 40-foot Collection Wall.

Most importantly, the coordination early in the design phase eliminated any change orders or changes to the space. As a result, infrastructure was installed in the right locations and there was always understanding of the technology that was going to be installed in the space. The locations of power, data, and technology infrastructure were planned for easy access for ongoing maintenance. This was essential for the Collection Wall. Design staff was able to make it architecturally elegant and maintenance-friendly without compromising the design of The Gallery One project team did benchmarking of other museums and exhibits that used technology, and talked to the people who maintain the spaces on a day-to-day basis to discover what worked well and what was problematic; that knowledge made the design of Gallery One better. By participating in the design phase for the physical space and art installations, CMA was able to create the software specification with a full understanding of the equipment that would be used and the technical support necessary.

The final result was a combination of hardware and software that work very well together and the Modular design was important so that spare parts can be on site and all exhibits can be repaired on site in less than an hour. The hardware was also designed early on so that a software or hardware failure could be mitigated by having the exhibit perform in a limited way. Even if the actual interactive is not working, there can be engaging signage on the screen rather than a dark monitor. Remote IP-based power switches are also integrated into all of the devices to allow for remote reboot of any exhibit to restore it in the event of a software problem that requires reset of the touch interface or display.


The Digital Asset Management Farm

CMA had been using Piction digital asset management systems for five years, and they multiplied during that time to three integrated systems. The main Piction DAM houses all of the CMA art photography for staff use, editorial photography, and soon, selected conservation and archival images. That DAM is internal, and links to the collections management system. It also feeds

  1. The DAM that supports Collection Online and serves up additional content for the website like epublications and lesson plans, and
  2. The CMS that provides artwork descriptions and images to the Collection Wall in Gallery One and ArtLens, along with rich-media and thematic tours, and over 850 videos.

Tying the museum’s DAM systems together has been challenging because the original data architecture of the collection records were based on the artworks instead of images. If an artwork had detail views—and some of them now have hundreds—it was almost impossible to manipulate the detail view images independently, to correct load errors, generate new derivatives, control use privileges, use in lightboxes, or export.

The Gallery One CMS

The images and data used in the Collection Wall and ArtLens are managed in a custom-built Piction CMS, built on Oracle 11g running on a Windows 2008 server. Object-related metadata is refreshed weekly from the museum’s collection management system in Piction DAM installations that support Collections Online and ArtLens, to ensure currency and synchronicity. Images for each object are transferred from the museum’s primary Piction DAM and post-processed to provide the 1.2 million image derivatives used to display artworks. The “cascading CMS” approach allows the Collection Wall and ArtLens to reflect current gallery installations (e.g., the art shown “moves” as it is moved within the museum, and “drops” when the object goes on tour/loan). Additionally, the Piction CMS allows hands-on management of over 500 video assets, 200 interpretive text assets, and more than 30 predefined tours, which can adjust over time to provide more and better interpretation to visitors. The Piction CMS runs on redundant hardware with solid-state drives to provide the best possible access performance for on-site visitors; a synchronized Amazon-based CDN provides images and videos for ArtLens users off site.

Digitizing the Collection

In 1996 the museum’s strategic plan outlined a commitment to becoming a national leader in the use of new and emerging technologies. A collections management database and digital scanners were purchased. By 1998, CMA was able to submit 1000 digital images with artwork metadata to the AMICO project. In 2003 CMA launched a massive expansion and renovation project which became an even greater force behind CMA’s image capture and inventory control. A dynamic inventory management module within the catalog database became a necessity, as the building project required the entire collection to be moved several times. At the same time, the initiatives to create digital assets for the collection for use in Collections Online received highest priority, as the administration saw the website as a way to continue to share the collection with visitors. As the collection was deinstalled and moved to storage, photographers were able to efficiently schedule large digital capture projects that would not be possible if the works were on view. Three photographers divided up the photography by type so that continuous steady progress is made across collections. CMA is well over 75% in its digital capture initiative and will have 100% digital capture in late 2014. High-resolution digital cameras that range from 48 to 192 megapixels were used to photograph CMA’s objects. These provide museum catalog-quality photographs as large as 50 by 40 inches, and will enlarge on a standard iPad or computer monitor to 220 by 160 inches for examination of detail.

Working with Artwork metadata

Optimizing Metadata for Collection Discovery CMA’s Apelles catalog is the primary source for all artwork metadata viewable on the Collections Wall and ArtLens. The Apelles database is comprised of approximately 100 tables, which feed 40 custom Views and two dozen Stored Procedures that allow it to feed dynamic, up-to-date collection catalog data to the Collection Wall. (Apelles is a 15-year-old, home-grown legacy Collections Management System that will be replaced with a new Collections Catalog and Management System in early 2015.)

When Gallery One opened in 2012, the Collection Wall fields for Collection, Date, and Medium were only mapped to catalog values for works currently on view, and the facets were set in a ‘one time’ update. During 2013 additional galleries were opened, gallery rotations kicked off, loans returned, and new accessions were added. This means new sets of catalog records became available for viewing on the Collection Wall weekly, simply by being moved into a public space via movement transactions in Apelles. IMTS staff quickly realized:

  • Utilities needed to run more frequently. New records flowed into the Gallery One database more quickly than anticipated.
  • Facet mapping tables needed to include the entire catalog, not just works on view, so that when artworks rotated on view they would immediately display in coverflows. This made the mapping analysis and assignment a bit more challenging, because the rule remained that each term needed to be represented by 12 works on view at any given time.
  • The creation date in the Gallery One database, which had originally been a one-time upload from a spreadsheet, needed to be added to the weekly refresh from Apelles.
  • A reporting and resolution process for catalog, image or hardware issues visible on the Collection Wall needed to be set up for the Gallery One frontline staff.

Incomplete Mapping Tables

The first issue to be resolved was the incomplete mapping tables. Combining and cross-checking search results across Apelles, Piction DAM, and the Gallery One CMS, the IMTS Applications Team was able to quickly identify the medium description strings coming out of Apelles, which had no matching term on the Material mapping table. Foreign characters in the medium description were causing the mapping update to fail, and analysis related to this issue provided a lot of opportunities to scrub data in the Apelles catalog, and the IMTS Applications Team worked with Collections Management to cleanup and standardize data. Once the new Material mapping table was in place and the utility ran, the project team saw a great improvement in the variety of artworks in the coverflows.

Creating Dates for the Collection Wall’s Coverflows

Next up was sorting out the Creation date. The Applications Team needed to find a way to include a sortable date in the Apelles to Piction DAM data flow without reprogramming the complex stored procedures. Analysis of CDWA lite script showed this data existed in Table 2, and further analysis of the data in Piction DAM revealed that the ingest already parsed out the field as CDWA.EarliestDate. Piction simply added this field to the DMZ and Gallery One data migrations. The date mapping table needed to be updated to change the formatting for ‘BC’ dates from the original { “-“ + date } to { date + “BC” }  – which is how the CDWA lite schema specifies the qualifier – and, as with the medium mapping, expanded to include all dates in the collection, not just those on view. This also greatly improved the breadth of artworks in the coverflows and improved the timeline view of the full collections on the Workflow for dynamic update of the Collection Wall facet: in this example, XML formatted ‘Creation Date’ is parsed, mapped to a generalized value having >12 and <25 representative artworks, and finally written into Materials facet the G1 database for use in the Collections Wall coverflow as well as related artworks functionality of the ArtLens iPhone app.

Once CMA resolved the data issues, Piction added the mapping utilities to the data migrations, and also delivered them our local Gallery One Administration tools so they can be run ad-hoc as needed.

Using the metadata for artwork made errors and inconsistencies easy to spot; likewise, actually using images of artwork in the ArtLens app and on the Collection Wall—which had been filed away in our Piction DAMs over the past five years—turned up more than a handful of digital images that were mistakenly associated with the wrong artwork before ingest, corrupted files, and photography for multi-item or multi-part artworks that was inconsistent in staging/lighting. Collections Management, the Photo Studio, and IMTS worked together to review and correct images as each new wing opened and additional artwork

A designated Gallery One help line (email account) was set up for Gallery One technical staff to report issues related to either catalog data or image appearance. Photos or videos of anomalies are attached to emails, which are then analyzed by the Applications Management staff to determine the root cause. Some issues are quick cataloging fixes, other requires restaging and photography before ingest, and others indicate issues with the mapping tables. All these issues can be quickly resolved in-house, as IMTS staff has control over all the underlying databases. Hardware integration or application software issues moved into issue resolution queue with the vendors.

The Beacon

The Beacon is a 4-by-4 array of 55-inch Edgelit 1080p LED displays. It plays a looping, non-interactive program displaying both dynamic and Content for “Dynamic” elements is pulled from the Sculpture lens. Individual face pairs from visitors playing the Make a Face game on the Sculpture lens are assembled into photo strips, with each photo strip containing four face pairs. New visitor photo content is loaded in at the start of each loop, approximately every six minutes. Content for top favorites is pulled from the network API. New content is loaded in at the start of each stage. The number of favorites is aggregated continuously over time, and the Beacon checks for updates to the favorites database. The content for the Beacon is generated and displayed in C++ and openGL, using the Cinder library. The entire loop is approximately six minutes long and cycles continuously. The events within the sequence incorporate the following: “playback” stages display only pre-rendered video; “rendered” stages display a composite of pre-rendered and generatively drawn elements; “dynamic” stages display content being pulled and rendered in real time. The Beacon deploys a video and communication bus to maintain uniform color across the entire display. It is driven by a remotely located Windows 7 workstation and extends video over Shielded Cat6 cable with digital extenders.

Wayfinding and WiFi

The museum’s IMTS department looked at a few different solutions for indoor wayfinding and selected the Navizon system, as it offered the greatest flexibility to be retrofitted into an existing environment. The Navizon system uses a series of small nodes that create a meshed environment that allows the system to know where a visitor is located. Once the nodes were installed and calibrated in the galleries, our application developer could tap into the Navizon API so that its location data could be incorporated into theArtLens app. CMA deployed Navizon’s Indoor Triangulation System (ITS): Over 100 Navizon ITS nodes were placed throughout the museum to locate, in real time, the iPads as visitors carry them through the museum. Even though the app was originally designed for iPad only, CMA chose this ITS because it tracks active Wi-Fi devices including Android, iPad, iPhone, laptops, and Wi-Fi tags with an accuracy of 2 or 3 meters, pinpointing floor and room. Though no application is actually required on the devices tracked by ITS, mobile apps like ArtLens can leverage Navizon’s API and be aware of the device’s position anywhere throughout the monitored area. In addition, the ITS will enable CMA to measure Gallery One’s success via indicators such as dwell time versus time interacting with the iPad — i.e., the museum knows how much time a visitor spends in a gallery and when they are interacting with the iPad Wi-Fi performance is key to successful user experience. The physical architecture of the CMA building and prior Wi-Fi configurations that were not sufficiently robust and posed challenges to the IPad performance and retrofit solutions that would be aesthetically pleasing were challenges that needed to be addressed. CMA selected a cloud-based solution from Navizon to provide location services in galleries for the ArtLens application. The Navizon system utilizes a series of wireless devices that link together using a wireless mesh which then relays the visitor location to a cloud base system, which can then be accessed via Navion’s API’s in the ArtLens application. Since the Navizon utilizes a small form factor device, these can be easily retrofitted into the galleries. CMA primarily uses a light- track fixture housing to contain and power the location device. While this was the preferred solution, it was not as cost effective as working with Navizon to bring the solution in-house. In early 2014, the CMA wireless network was migrated from an autonomous base wireless network to a centrally managed controller based wireless network. This allows CMA to gain greater visibility into the overall performance of our wireless network and make changes dynamically based on current usage.

In addition, all wireless access points in galleries were recently upgraded to provide additional speed and capacity using a Cisco 1262 access point. This allows CMA to provide speeds up to 100mb (802.11n) as well as using both 2.4 GHz and 5 GHz frequency bandwidths for the most optimal performance on the visitor’s mobile

Upgrading Wi-Fi and Wayfinding in the Galleries

The Cleveland Museum of Art (CMA) faced many challenges during the design and actual build out of the wireless network. The first challenge was that the initial design process took place back in 2003 at a time prior the wireless access becoming as widespread and expected as it is today. The second challenge was to install the devices in galleries in a way that was aesthetically appropriate to a gallery space. This meant that all wireless hardware had to be mounted above the ceilings in locations where access hatches were placed based on architectural design, which did not necessarily work for a proper wireless network design.

This location issue became more of an issue in 2012 when CMA was planning its ArtLens application, which relied on using a wireless network to obtain the visitors position so that the content appropriate to the area could be automatically delivered.

CMA’s initial thought process to achieve location-based services was to leverage the existing Cisco wireless network that was already installed. CMA engaged its wireless vendor to determine what would be involved to provide location-based services in our environment. Our vendor analyzed our existing installed wireless network to determine the number of extra wireless access points that would need to be installed and where they would need to be located so that the ArtLens application could properly triangulate the visitor’s location within a particular gallery.

Recommendations were made by our wireless vendors (red circles) based on our current wireless network. Installing these new access points would involve cutting in access hatches in the ceilings and pulling CAT6 cabling to each new location. Since the gallery spaces were already built and installed, this would involve closing and deinstalling galleries which CMA management determined was not feasible.

Since location services were still needed, CMA had to look into other options which could easily be retrofitted into the galleries. After doing additional research, CMA selected a cloud based solution from Navizon to provide location services in galleries for the ArtLens application. The Navizon system utilizes a series of small form factor wireless devices that link together using a wireless mesh which then relays the visitor location to a cloud-based system which can then be accessed via Navion’s API’s in the ArtLens application.

Since the Navizon utilizes a small form factor device these can be easily retrofitted into the galleries. CMA primarily uses a light-track fixture housing to contain and power the location device. NOTE: A white cover is attached to the bottom of each fixture so that only the location device’s antenna is visible. The gallery ceilings at CMA have light track fixtures running through out which makes the installation and placement of these location devices very flexible. CMA also was able to install the location devices in display cases that contained power. Below is an example of how the location nodes are positioned throughout the CMA galleries to provide the best possible triangulation location information to the ArtLens application.

As the usage of mobile Wi-Fi enabled increased, CMA found that the wireless coverage in the galleries was becoming inadequate with respect to speed and connectivity. Over the past year the CMA wireless network was migrated from an autonomous base wireless network to a centrally managed controller based wireless network. This allows CMA to gain greater visibility into the overall performance of our wireless network and make changes dynamically based on current usage. In addition, all wireless access points in galleries were recently upgraded to provide additional speed and capacity using a Cisco 1262 access point. This allows CMA to provide speeds up to 100mb (802.11n) as well as using both 2.4 GHz and 5 GHz frequency bandwidths to provide the most optimal performance to the visitor’s CMA has also implemented best practices recommendations from Cisco for using iOS-based devices: bestpractices/EntBP-AppMobDevs-on-Wlans.pdf

Given the architectural challenges faced by CMA’s historic 1916 building, with its thick metal lath and plaster walls, the wireless signals do not propagate as they would in a more modernly build building. This has become more of an issue with mobile devices which have more limited capabilities to pick up wireless (Wi-Fi) signals. CMA has worked with our wireless vendors to do a survey study as to how our current wireless network infrastructure would be seen by the mobile devices most commonly used today. The study has revealed areas where coverage would be inadequate as shown below and actual field testing has confirmed these results.

CMA is currently in the process of adding additional wireless access points in galleries to help supplement the amount of Wi-Fi coverage available. Even with all of the various upgrades, additions, and configuration changes to our wireless network, IMTS has still found the client devices (specifically iPads) to not perform as they should on an enterprise class wireless network (such as dropping their connections and not properly roaming). CMA is not alone in this challenge. In working with multiple wireless vendors, CMA learned that the Apple products do not always “play” as nicely as they should with enterprise-class wireless networks. In addition, Apple keeps their wireless hardware and software very well guarded which makes the necessary testing process to debug very difficult. To date, IMTS has tested several different wireless configuration settings to overcome these challenges in addition to reviewing how the ArtLens app itself is utilizing the wireless network for both location-based and data transfer functionalities.

NOTE: Even though Apple indicates they support Fast Roaming (802.11k) for iOS based devices, CMA found out through testing and confirmation from its wireless consultants that it does not currently work. ( IMTS is currently working to address our location based system in our new North and West wing galleries, which have walls that do not go from floor to ceiling and also have multiple galleries in an open area (such as a corridor). In this situation, the Near You Now location indicator in ArtLens bounces around between galleries depending on the user’s location in one of these open wall galleries. IMTS is also working with Navizon to review the placement of the location nodes in these galleries to determine optimal placement to help address this issue.

The other challenge the museum is currently working to address with the location nodes is how best to avoid the wireless channel overlap between Navizon and Cisco networks. Since both systems utilize the 2.4 GHz wireless frequency, and specifically because of the Navizon’s nodes use of Channels 11, IMTS had to turn off the use of Channel 11 on the Cisco side to avoid performance issues between the two systems. This is causing high utilization on the Cisco network for devices using the remaining two non-overlapping 2.4 GHz channels. This is most noticeable in areas that have a denser group of devices on the 2.4 Ghz wireless network. CMA is currently in the process of moving our wireless devices to the 5 GHz wireless network to avoid this congestion. Navizon has also indicated the nodes they currently use will soon be able to work on the 5 GHz arena.

Wait, what about iBeacon?

iBeacon is a proximity-based system that only allows for a very rough estimation of the nearness of a device (specifically far/near/immediate). So it can’t be used to determine absolute position; the museum’s Navizon-based Wi-Fi trilateration system can. A proximity-based system like iBeacon could be used in place of RFID to push notifications from the Collection Wall docks to mobile devices with ArtLens, including Android. If the device detects an “immediate” beacon, it’s docked.

Digital Sharing: Twitter and Facebook

One of the most complicated back-end re-alignments necessitated by ArtLens was a complete rebuild of the museum’s public website and Collection Online to provide constructible page addresses for artwork. The museum wanted visitors to be able to share their favorites via social media through ArtLens, but original website implementation offered its content pages in popups, and hid nonsensical, utilitarian URLs from online visitors. There was no way to get from one to the other easily, and no way for Google’s crawlers to ‘see’ our artwork! IMTS migrated the museum website to Drupal in December 2012 to support linking straight to a specific page, specifically to construct ‘real’ URLs for artwork pages (based on accession number) dynamically.

Favoriting and sharing activities on ArtLens drive significant traffic directly to the web pages for artwork on view, completely out of context of the ‘search the collection’ page. IMTS worked with a Collections-focused team to redesign and enhance the artwork pages for better usability. The team is now reviewing process (and permissions) to include ArtLens assets for artworks which have gone off-view. The Drupal CMS for Collection Online online is managed in-house which provides the team with the necessary flexibility and control to implement this expansion to the data flow.

With the museum’s rental iPads, additional log-in prompts for social networks must occur because they are not visitor-owned and do not store previous registration.  Initial analytics after the first launch of ArtLens showed minimal use for social sharing, at only 4% of users. This led us to implement the broader share capabilities of email and texting for the iPhone and Android versions. With the iPhone version, more visitors are sharing because with iOS 7, registration and log-ins for Facebook and Twitter are automatic and appear within the standard iOS share sheet. The visitor’s email client also appears and when artworks are shared, an email is sent from their own account that contains CMA crafted default copy, subject line text and layouts. With this method, a small increase in sharing has been noted. Time and more analysis of usage statistics will reveal more details on which method is utilized the most for this feature. At this time, the rates are fairly even across social media and email sharing.

Post-launch: Improvements, Preventative Maintenance, and Troubleshooting

To make sure the technology enhances the artwork and does not distract from it, the CMA laid down a number of ground rules for the technology it employs. For example, while many of the effects appear cutting-edge, the technologies underpinning them are structured with best practices in mind, and thus less likely to fail. Similarly, CMA’s team keeps spare parts on site so that any system can be repaired within 45 minutes. If software fails, then devices can be rebooted remotely via an IP-based power switch. Inactive screens still show content, so visitors are never confronted with a blank display. Maintenance for the Gallery One Interactives has focused on preventative maintenance that includes regular cleaning of the electronics including screens, sensors, fans and filters. The technology design of Gallery One is focused on sustainability and reliability. During the many benchmarking visits the Design team conducted to other museums, a repeated theme was equipment failures and downtime. It was critical to the Design team that Gallery One experience minimal downtime through robust equipment and being able to completely replace any component within 60 minutes of a technician being on-site. To make this a reality, the Lens housings have been designed as interchangeable modules that separate the display, interactive and control components. Wherever possible, common components have been selected and spare parts for each are stored

The Collection Wall and ‘Line and Shape’ interactives use the latest in modular technology, and this is the key to the sustainable design. Each of the 162 video tiles in the Collection and ‘Line and Shape’ walls is a modular 16”x12” cube which can be completely rebuilt from the front without removing the chassis from the wall. Complete replacement of a tile’s internal components can be completed in 30 minutes. The Touch Interactive technology is so new that the exhibit was initially opened with beta hardware from the manufacturer.  The AV integrator, Zenith Systems, worked closely with CMA and Christie Digital through the development of the touch interface so that Gallery One was the first installation to take advantage of the new technology. The touch interface is designed in standard 12” and 16” sections which can easily be replaced in the event of a hardware failure. This approach maximizes the “up-time” of this signature exhibit through a sustainable design.

The head-end systems, which include the server and extenders, are inspected to maintain optimal temperatures and the stability of media storage. Spare parts for all systems are stored on site to reduce down time and expedite repairs. Zenith Systems has developed software to run on the servers to test the function of the displays and touch interactives and audio. When an issue is reported, the first step of diagnostics is to shut-down the exhibit software and test the complete software path with the diagnostic software. This is good software with reporting on touch responses that are visual to the technician. It is critical to the support of a complex interactive to have a method of troubleshooting that reliably identifies the difference between a hardware malfunction and software application issue.