MUSEVUE

Inspiration

We were curious about how museums could leverage data to improve visitors’ experience. Museums and galleries are important institutions for cultural education, both for citizens as well as tourists. Over the years, museums have seen a decline in attendance, especially in an increasingly competitive digital space.

Thus, it is vital that the museum experience is not marred by distractions such as “hyper-congestion”, which can greatly reduce the amount of time spent in a museum and the quality of visitor experiences. Our goal is to streamline a visitor’s experience, making the art that they do encounter more impactful.

What it does

What we’ve seen is that many museums are not tracking this data at such a granular level. For one thing, it is incredibly cost-preventative and only museums with as much foot traffic as the Louvre would find bluetooth sensor installation necessary.

Our website is a real-time map that advises visitors on which galleries/exhibitions are best to visit based on foot traffic. Upon entering a museum, visitors will be prompted to go to the website to see a map of the gallery as well as a “heat map” of sorts that will indicate the most populous galleries, which would deter some visitors from visiting those galleries. Ultimately, this will be able to control some of the hyper-congestion museums face as well as manage flow.

How we built it

First, we simulated a gallery by taking intermittent images of a room of people and utilized Google Cloud’s Vision API to count the number of people.

There is a lack of public museum data, so we simulated foot traffic based off of data from Disney Crowd Calendar for our demo. We used Python and Django to calculate an integer output. Our algorithm considers data points such as visitor preference, relevance to main exhibit, crowdedness, as well as exhibit capacity to determine the best exhibit for the visitor to visit. Thus, the experience is custom-fitted for every visitor.

Finally, we developed front end using HTML and CSS to build the website via Adobe Dreamweaver.

Challenges we ran into

We encountered many problems because we did not have a lot of experience in creating a web app. It was our first time using Google APIs, and it didn’t work the way we wanted to. Additionally, we struggled with learning framework such as Django or BaaS such as Firebase to service our site. Finally, we had a very basic understanding of HTML/CSS at the outset of this project, and it required practice with Adobe Dreamweaver to understand how the code works.

Accomplishments that we're proud of

We’re proud that we figured out how to adjust the Google Vision face detection API to fit our needs. We’re also proud of the algorithm we used to calculate the output in Python. Additionally, we have a basic understanding of how Django works. We were able to improve our front-end development skills. Ultimately, we’re proud that we made a website that works! Even though it wasn’t exactly what we had in mind, we now know for any of our future projects.

What we learned

We learned how to work as a group. We distributed roles based on our backgrounds and did our best to solve the problems that we encountered. We learned more about the website development process through trial and error, and we have a better understanding of timelines.

What's next for MUSEVUE

In the future, we would like to work on a more interactive map that lights up rooms based on it’s capacity (from green to red). It would look more like a heat map and visitors would have a more visual understanding of capacity. We also would like to flesh out our “explore” section, which would suggest exhibits to visit based on user preferences and location data (if given). We also think a “forecast” section would be relevant for visitors who are planning their trips as tourists--similar to Disney’s Crowd Calendar but on a more granular level. Although museums usually have estimated figures for attendance, we’d like to make this data publicly available so that exhibit space is more efficiently visited.

We think this technology could be relevant in many other spaces. Although it is already used in retail spaces with sensors, our technology is a more cost-effective option for public areas. One implementation of this could be for bathrooms in locations with a lot of traffic, such as malls or parks. By managing the flow of visitors in public bathrooms, we would be able to redistribute visitors more effectively and reduce queueing times.