Here’s What a Smart Museum Could Look Like

Reinventing museums through IoT could transform the way we experience art. It could also inspire entirely new business models for future museums. Are you ready to visit a smart museum?

Smart Museum — IoTforAll

The rapid rise of technology in real estate brings along a stream of valuable new opportunities to building owners. The results from the first wave of pilots are now demonstrating the value of upcoming technologies in optimizing building performance, enhancing the user experience, and realizing cost efficiencies. Widespread adoption is about to take off.

Up until now, use cases have mainly revolved around commercial real estate, with shopping centers and office buildings leading the way. The use of new technologies within non-commercial (or semi-commercial) real estate, however, still is a virtually unexplored territory. With technologies becoming more and more accessible by the day, both technically and financially, it is time to start exploring how these could add value to the next frontier of smart building adopters.

To illustrate this exploration and the possible value that lies within, I will describe how making a museum ‘smart’ could positively enhance a museum’s way of conducting business, boost visitor experience, and perhaps even increase its impact on society.

First, I will start by briefly describing how museums generally handle business. It is important to know the workings of any company before one can identify how the application of technology can make a meaningful impact; we want to go beyond gimmicks.

Next, I will describe how increasingly advanced technology can be applied to unlock new opportunities for museum owners and their visitors. Among the technologies that I will discuss are the Internet of Things (IoT), facial recognition cameras, and biometric sensors. I will conclude the article by illustrating, through an example, how new technology can lead to entirely new business models in the art industry.

How Museums Currently Do Business

Before proceeding to explore the technologies that may prove valuable when adopted by a museum, we first need to gain an overall understanding of how museums usually operate. When we look at the key performance indicators* of a museum the two major ones are: the annual number of visitors and its annual revenues, where attracting the visitors leads to the generation of the revenues. So attracting visitors is crucial for a museum in sustaining itself. But how does a museum attract visitors?

This is where the museum’s collection comes into play. The collection embodies the museum’s core value proposition for attracting visitors: “If you come to visit us, then we will show you interesting objects [x], let you experience [y] and teach you about [z]”.

From a business perspective, one can say that the permanent collection generally contains the high-end pieces to ensure a certain degree of quality and attractiveness throughout the year. The rotation of temporary exhibitions makes the museum more attractive for visitors to return regularly. This combination should create a sustainable stream of income for the museum.

Besides its primary revenue stream from ticket sales, secondary revenue streams come from the museum’s restaurants, souvenir shops, and special events. Note that the first two of these secondary revenue streams are mainly driven by the museum’s number of visitors, and thus the collection. The occasional sale of a piece of art may also yield revenues, but repeating this too often cannot be considered a sustainable way of doing business.

So, a core asset for a museum to sustain itself is its collection. This makes the collection a perfect subject for further exploration. How can technology contribute to creating the ultimate collection with the purpose of enhancing the museum’s business?

In the remainder of this article, we will explore a variety of methods to answer this question.

*Factors such as the type of collection that the museum carries, whether it is publicly or privately owned, for-profit or non-profit, funded or non-funded, are of less importance here.

Image Credit: Chris Karidis (Unsplash)

IoT Applications in Museums

Now that we have established the importance of a museum’s collection, we will proceed to explore how a museum can harness technology to enhance its offerings by:

  • Gaining insight into the target audience (visitors) by collecting data about their behaviors and interests.
  • Finding new ways to measure a collection’s performance, moving beyond the number of visitors and revenues.
  • Tailoring future exposition offerings to the interests of (potential) visitors.

In this section, I will describe the application of new technology through five consecutive stages. The first stage starts off relatively basic, and one may assume that some museums have already started this step. In the steps that follow, however, the applied technology becomes increasingly advanced.

This advanced technology will generate more detailed and accurate data, and potentially has a significant impact on the smart museum’s business performance. The applied technology at the same time brings along new metrics that are likely to be used to measure the ‘performance of art’. These metrics will be introduced along the five stages.

Meet Metric #1: ‘Art Attention Score’

“Tell me to what you pay attention and I will tell you who you are.” 
— Jose Ortega y Gasset

In the stages mentioned below, we will use technology to rate the performance of an individual piece of art based on the level of attention that it manages to capture from museum visitors. This first metric is based on the assumption that a visitor visits an exhibition with a certain curiosity and/or expectation. The idea is that the longer a piece of art holds the visitor’s attention, the more it meets (or contradicts) this curiosity or expectation, and the more important the piece of art is to the collection.

Stage 1: Tracking Visitors’ Journeys

In the first stage, we start by installing a people counter by the entrance of a space. This allows us to track the number of visitors that enter and exit each space within a museum. From analyzing the generated data from the spaces, the smart museum owner can see which spaces are most frequently visited by patrons, and which art pieces thus attract or generate the most attention.

For example, a report from the Louvre museum in Paris may show: “90% of today’s visitors have entered Space A, which contains the Mona Lisa, while 25% of the visitors that day have entered Space B, which contains pieces from upcoming local artists.”

While the report above may come as no surprise, the data would become particularly interesting if two spaces were to hold pieces of similar “quality” but have data showing a significant difference in the number of visitors that have entered the spaces. Is this really because the visitors find the art inside one space more interesting than in the other? Another reason for it could be that wayfinding or signage is off. What would happen if these were tweaked? Or what would happen if we would swap the pieces between the spaces?

The data collected could drive experimentation and show the impact of possible improvements. The major limitation of this approach, however, is that it would be impossible for us to identify how much an individual piece of art contributes to tempting visitors to visit a particular room. Each piece would need its own unique space with its own people to measure the ‘attraction’ of each piece accurately, which would be far from an ideal. Even big museums like the Louvre would quickly run out of rooms. Fortunately, advanced technology can solve this issue.

Image Credit: Riccardo Bresciani (Pexels)

Stage 2. Real-Time Visitor Location Tracking

In the second stage, we track the location of visitors within each space through sensors installed in the ceiling. This allows us to accurately pinpoint where visitors are standing and to analyze their movement patterns throughout the space.

In general, when a visitor inspects a piece of art, he or she stands within a standard distance from the piece; closer to smaller pieces and farther away from larger pieces. For every respective piece, however, this distance remains fairly constant. From this, one could assume that whenever a visitor stands within this predetermined distance from a piece, he or she is paying attention to it. Let’s call this the ‘attention spot.’ This attention spot can be used to trigger a scoring mechanism: whenever a visitor enters the attention spot of a piece of art, it will yield attention points for this piece of art — allowing us to measure its degree of ‘attractiveness’ in comparison to the other pieces in the space and the smart museum.

For this to work, the art pieces need to be positioned around the space so that their attention spots do not overlap. By doing so, we actually subdivide the physical space into virtual smaller spaces, allowing us to measure the attention per piece of art — not per space — solving our problem with stage 1.

Furthermore, the attention spot could be used to generate data in a variety of ways:

  • One attention point is earned by the piece of art whenever a visitor walks through its attention spot.
  • Two points whenever a visitor walks through the attention spot very slowly.
  • Three points whenever a visitor stands still for one second, and more points the longer he or she stands still.
  • Four points whenever a visitor walks up for closer inspection.
  • Etcetera.

Using this technology, we can still only assume that visitors are paying attention to a piece whenever they are in its attention spot. A visitor could also be talking to a companion or be looking at his/her phone.

The application of a different and more advanced technology, however, would allow us to measure the attention that is given to each piece of art far more accurately. So let’s move on to the next stage, while at the same time adding another metric to the mix.

Meet Metric #2: ‘Art Emotion Score’

“Who sees the human face correctly: the photographer, the mirror, or the painter?” — Pablo Picasso

Now that we have taken a first step in exploring attention as a new metric to rate the ‘performance’ of art, it is time to juice things up. For the next stages, we will attempt to use the latest technology to see whether we can take a first step in measuring the emotional impact of a piece of art: the ‘Art Emotion Score.’

Image Credit: Lafon Pauline (Twitter)

Stage 3: Facial Recognition Cameras

In this stage, we no longer use a visitor’s location to guess whether he or she is paying attention to a piece of art but apply facial recognition technology to be sure of this. We do this by installing a camera above each piece, capturing the visitor’s face and analyzing it. That would allow us to collect data on:

  • Whether and how long an individual visitor is paying attention to the piece of art (and even to which part of it).
  • The visitor’s demographics: age, sex, height, weight, ethnicity, etc.
  • The visitor’s emotions: joy, disgust, surprise, discomfort, horror, etc.
  • The combination of all viewed pieces of art in the smart museum by the visitor, creating unique visitor preferences profiles.

Applying this technology opens up a treasure trove of new insights and potential value. For example, it would allow us to allocate emotion points to art — along with attention points. What percentage of visitors smiled while looking at a painting? Or looked sad? Or angry?

This emotional impact could be linked to the goals of a smart museum. If a museum’s goal were to educate its visitors on the negative impact of war on local societies, then a target performance metric may be to have at least 70 percent of the visitors show emotions of discomfort three times per visit, provided users don’t find a way to fool facial recognition systems. At the same time, this information about emotional impact may prove highly valuable to artists. What is the emotional impact of the piece they created? Was the emotional impact as intended?

Data from this report illustrates: “The Mona Lisa performs very well in holding the attention of women between 30 and 44 years old, especially from the southern part of Europe. At the same time, the painting shows an above-average emotional impact on their ‘happy’ emotion.” The smart museum may decide to advertise in Southern Europe, acting upon the report to attract more visitors. They might show a picture of Mona Lisa in a ladies magazine, for example.

Stage 4: Visitor Identification

In this stage, we use facial recognition technology to track all metrics as described in the previous stage, but add a social layer to it. Instead of merely recognizing a face and tracking where this face reappears throughout the museum during a visit, we also connect the face to the actual identity of the visitor by using social media. And this is what some might find unsettling.

So how would we go about this? It’s much simpler than you think. Facebook already knows what you look like from analyzing the unique facial structure from your Facebook photos, and it obviously already knows your name. At the same time, the tech giant is investing heavily in technology that will enable it to identify you from any (live) video source using facial recognition combined with the facial structure data already present in their database.

From using this newly acquired data, a museum report can now read: “Daan de Geus from Amsterdam smiled while paying attention to the Mona Lisa for 113.6 seconds. This attention span is 34% longer than the average for his visitor archetype. From analyzing the preferences of Daan’s 481 Facebook friends (through their ‘likes’), we identified that 16 of Daan’s friends might find the Mona Lisa interesting. Two of these friends have already shared pictures taken inside the Louvre last month. I have therefore automatically paid Facebook to advertise the Mona Lisa on the social media profiles of Daan’s relevant 14 friends.”

Stage 5: Biometric Sensors

The fifth and final stage is a highly experimental one. In this stage, we move from collecting smart museum visitors’ behavioral data to collecting data from the insides and surfaces of their bodies. Our goal is to detect changes inside the body, caused by emotions that may be triggered by a piece of art. We could measure:

  • Heartbeat. Here we measure the pulse of visitors through pulse sensors from a distance, allowing us to measure (the average) change in heartbeat per space or per piece of art.
  • Temperature. Certain emotions have shown to lead to changes in temperature across specific areas of the face and body. By adding thermal infrared to our facial recognition cameras, we can pick up on changes in emotional state at a deeper level.
  • Tone of Voice. Microphones can be installed to allow us to analyze the tone of voice of the visitor. An above-average high-pitched tone of voice around a piece of art can be an indicator of a specific emotion. Note that this does not track what is said, just how it is said.

A smart museum report could read: “While viewing the Mona Lisa, the average change in body temperature per visitor was +0.6 degrees Celsius. At the same time, a significant average increase in blood flow to the forehead and the tip of the nose was detected, possibly indicating the emotion of joy. Heartbeat increased by an average of 1.4 BPM while tone of voice identified the emotion of ‘excitement’ on a level far higher than the average across all spaces.

Image Credit: Berkovitz et al. (2014) from Psychophysiology

From Data Insights to Real Business Opportunities

“Small opportunities are often the beginning of great enterprises.” 
— Demosthenes

Through the application of technology presented in the five stages above, a wide range of valuable new data is collected. The example reports after each of the five stages illustrate how this data can directly lead to actionable insights. There are, however, many more opportunities for turning the newly generated data into value. Below, you will find eight:

  • Data triangulation. While the different methods of data collection throughout the five stages can be valuable on their own, they may at the same time be prone to bias. Are we measuring what we think we are measuring? By overlapping the data collection techniques across the stages, a much more accurate analysis can be made and new insights can be gathered.
  • Collection customization. From the data that is collected, the smart museum gains valuable insights into the preferences of its target market. Based on these insights, the museum can tweak its current collection and design its future exhibition offerings to attract more visitors.
  • Existing data sources. There are many more data sources available from inside and outside a museum that are waiting to be used. What if we linked outdoor weather conditions to emotions?
  • New meta metrics. In this article, we explored new metrics to rate the performance of art. Combining these metrics with other datasets may possibly lead to new meta metrics. For example, learning theory states that in order to learn something new, the student needs to be exposed to a variety of examples. Could this mean that when a visitor has paid attention to ‘y number’ of pieces for at least ‘x seconds’ within a theme, we can imply that the visitor has learned something?
  • Cross-museum benchmarking. The data that is generated within a smart museum can also be used to benchmark across museums. Which museum has the highest overall emotional or educational impact? How does the same piece of art perform across different smart museums?
  • Cross-sector application. The technologies presented in this article, and how they are applied, can be valuable to building owners in other sectors as well. Why would we not want to measure emotions in retail stores? How about measuring employee heartbeats in the workplace to monitor and reduce absence due to sickness or burn-out?
  • Visitor recommendation engine. When data is collected about a visitor, its value can be returned to the visitor as well. An example would be a recommendation engine: “Based on your preferences (attention and emotion scores) from museum A, or across museums B to D, you should definitely visit museum E.”
  • Smart art. As stated earlier, artists could use the newly generated data to analyze the emotional impact of their art. At the same time, they could also use the new technology to create interactive art. How about an abstract painting that responds in real-time to your heartbeat, viewing angle or facial expression?

New Business Models Through IoT Technology

Apart from the various opportunities that new technologies may bring, their application can also lead to entirely new ways of doing business for a smart museum. In the example below, I will illustrate how a small change in the business model of a museum could affect the whole value chain of art.

The ‘Pay-Per-View’ Smart Museum

Since facial recognition technology can register what pieces of art a visitor pays attention to, it would enable us to charge visitors only for the art they look at. In practice, this could mean that if a regular entry ticket would cost € 20, we would instead charge € 1 per view (what a ‘view’ exactly is, would need to be defined). If a visitor’s total number of views exceeds 20, then the visitor’s pay-per-view ticket would automatically switch to the regular ticket, so that he or she will never pay more than € 20.

A core benefit of this model is that it would attract visitors to the smart museum that only want to come in to see a few pieces, perhaps for studying or drawing but are put off by the high price of a regular ticket. A pay-per-view option could in this way bring more visitors into the smart museum, potentially boosting secondary income streams through restaurants and souvenirs. While inside, the smart museum could attempt to ‘seduce’ these pay-per-view visitors into viewing more pieces, resulting in a visitor that otherwise would not have come inside paying the full ticket price.

The adoption of this pay-per-view model could also impact other links in the value chain of art. It would allow a smart museum to rent a piece of art from another party and only pay for the number of views that the piece of art generates, or even the number of specific emotions that the piece generates (depending on the goal of the exhibition). At the same time, this model would allow the artists to earn royalties per view, leading to a model that is similar to that of YouTube and Spotify, where artists and publishers earn a predetermined sum per x-number of plays.

Image Credit: Pexels

Conclusion

Smart building technology is on the rise and the applications and potential seem endless. This article illustrates how using technology is not exclusive to highly commercial or technology-driven organizations. Every building owner can — and perhaps should — start to explore how smart building technology can be applied to enhance their business.

We’ve learned that new opportunities can come forth not only from the direct application of a new technology, but also from combining it with other technologies or with already existing and new datasets. Moreover, we see that the rise of technology can unlock alternative ways of doing business for an organization, potentially affecting the whole value chain in which it operates. Being the party taking the initiative to bring this change has proven to lead to unique positions of power. However, it’s also important to use IoT in a calculated way, choosing what to connect and what to leave offline.

Particularly for smart museums, the application of new technologies can potentially lead to both securing financial stability and social relevance for the future. With technology becoming increasingly sophisticated by the day, a smart museum’s previously unmeasurable impact (education, emotion, change in behavior) has started to become measurable.

How to get started

Do not wait to strike till the iron is hot; but make it hot by striking.
 — William Butler Yeats

As illustrated in this article, adopting new technologies within your building can lead to previously unattainable benefits, and a small first experiment can be conducted with ease. However, implementing new technologies within a building often comes with hurdles and perhaps even resistance. To manage these challenges effectively and to increase your chances of success, I highly recommend keeping the following guidelines in mind:

  • Start with designing an experiment that aims to solve a direct problem for your organization — and preferably for the end-user at the same time. This boosts the chances of acceptance from your most important stakeholders.
  • Create a small area inside your building in which you experiment with new technologies. Make sure that this area is used in the same way as the rest of your building. This increases the generalizability of the findings from your experiment throughout the building.
  • Communicate why the data is collected, how it is collected, how you secure it, and what you intend to do with it. If necessary, design an opt-in for your experiment.
  • Consider beginning with collecting data that cannot be directly linked to individuals. While this data may be less specific, it is also less privacy-sensitive and less complex in general, while still opening up plenty of opportunities for further exploration.
  • Give back the value to the end-user, for example through a simplified dashboard with insights. In this way, the end-user also benefits from participating in your experiment, strongly increasing acceptance.

Are you ready to smarten up?

Disclaimer: The application of technologies as presented in this article may have serious privacy implications. This article is all about exploring and illustrating the possibilities that arise from the application of new and upcoming technologies. Of course, any privacy issues that come up need to be fully addressed before implementation of any technological application takes place.

This article was originally published through IoT for All—the world’s hub for the Internet of Things.