This year’s Super Bowl halftime show opened with a shot of Lady Gaga seemingly performing on the roof of NRG Stadium in front of an impressive, patriotic and rather animated constellation — with a hankering for Pepsi. The light array, which was actually a squadron of 300 Intel “Shooting Star” drones, along with Gaga’s rooftop performance, provided the show’s pre-recorded wow factor before the pop star landed on stage for the live performance.
While the spectacle certainly takes its place in the pantheon of eye-popping halftime extravaganzas at the big game, to fully appreciate both the artistry and technology behind the show it might have helped to get a closer look at the swarm of sentinels. To help understand how these automated acrobats are working their way into stage performances, Youngmoo Kim, PhD, a professor in the College of Engineering and director of Drexel’s Expressive and Creative Interactive Technologies (ExCITe) Center, took some time to break it down — and assure Gaga’s backup dancers that their jobs are safe.
In a partnership with Parsons Dance and NextMove Dance, Kim has been advising a team of undergraduate students in engineering and computer science, who developed and programmed drones as part of an interactive dance performance that debuted in December. The drones, which were largely 3D printed and assembled by the students, are a bit larger than Intel’s squadron, but they are designed to share the stage with the dancers, rather than slinking into the background. The drones and dance companies will open up a multi-week run at New York’s Joyce Theater in the spring.
Aside from organizing and maintaining the sheer number of drones involved in the halftime performance — some 300 — what is the biggest challenge for pulling off a performance like that versus the one you created with Parsons and NextMove?
The biggest difference is that the halftime show was outdoors and ours is inside. When you’re programming drones for something like that, positioning is everything. They need to know where they are in relation to each other. When it’s outdoors, you can use the GPS to position them precisely where you want them.
But for an indoor performance, you can’t use the GPS because the signal is obstructed. So you need to come up with another way for the drones to get relative positioning information and respond to it. In our case we set up a motion capture system in the theater, so the drones could take their readings from that and adjust their positions.
How important are the control systems in performances like this? How did they differ for your indoor performance?
Well in our case, with a few drones on stage, very close to the performers and very close to the audience, it was very important to have numerous safety procedures in the event of a malfunction.
In Intel’s case, they invested a lot of money into the control systems to ensure that there was redundancy. So if there was a problem, one drone could seamlessly drop out while another was taking its place. That’s something you can do when they’re far away, but not when they’re right in front of you.
And there were plenty of other safety measures in place for the big performance, obviously the FAA had to limit when they could fly in that air space — which is part of the reason it had to be pre-recorded. So it’s understandable that they would invest a lot in reliability and redundancy.
How would you assess the drones’ performance at the Super Bowl vs. your drones in the Parsons/NextMove ensemble?
Well, the Intel drones weren’t really at the Super Bowl, so that was kind of a bummer. They had to pre-tape theirs, so who knows how many takes they needed.
We had to do it live and we had seven successful performances — without any crashes.
Not to take anything away from what they did, it’s a really cool big budget thing to throw into the spectacle. But when it first started happening I didn’t know that they were drones, it could have been CGI (computer-generated imagery). They had Gaga looking like she was on top of the stadium — and that was fake, so it’s easy to miss the fact that the drones were real, not just a visual effect imposed later.
It was an impressive technology demonstration, but ultimately no one there saw it live and no one saw it live on television…so what was the point? — Just to have a big ad for Pepsi in the end.
At a certain point, for it to have an impact from an artistic perspective you need to do it live, so the audience can see it.
How big a step is this for incorporating this sort of autonomous or semi-autonomous technology in artistic performances?
It’s a big step because people are thinking about how they can incorporate technology into artistic performances. But I don’t think it necessarily drew the audience in.
In our case, the people in the theaters could feel the wind being pushed around by the drones. People in the front of our audience could literally feel the drones. On television, you don’t get that same sensation.
The thing we try to do is make that connection between drone and audience more meaningful. Tech isn’t a solution in search for a problem. We decided to combine drones with dancers to explore and reveal what could happen with drones next to humans.
Putting 300 drones in the sky is impressive in its own way, but at that distance you really don’t appreciate the fact that they’re drones. Being able to see them close up, and get a better sense of scale by having them next to the dancers, makes people realize how big they are.
They actually feel the wind created by the propellers and it is a bit scary to have these things whizzing by your head. There is a sense of immediacy and it’s impactful in a different way than putting 300 drones in the sky — which is cool, but it’s not as interactive, it doesn’t allow people to connect with them.
How does it help to change the perception of what drones and UAVs can be used for? (not just surveillance, attacking enemies and as remote controlled toys…but part of artistic expression)
To a degree it helps, but it wasn’t necessarily obvious that those were drones. You could easily look at that part of the performance and think that it was all CGI. She wasn’t really jumping off the roof and those weren’t really 300 flying robots.
It’s a subtle point, but this is one of the reasons we try to do things live — so that people can see for themselves that it’s real, they are part of the performance.
So when there’s a show like this and it’s almost virtual in the way the drones are perceived by the audience, in some ways that actually increases the distance between drones and reality.
I know these drones were part of an installation at Disney that ran for a while, so depending on how close the audience could get to the drones there, that might have had more of an impact on people than the Super Bowl show.
What’s next? …Did seeing that show give you any new ideas?
Our next event is the official premiere of the piece in May with Parsons Dance in New York City. There will be 14 performances at the Joyce Theater. And our senior design teams will keep working to improve the system. It’s a two-drone system right now and we might look at adding more to it. Though we’ve built 10 drones so far, it’s hard to gauge what it’s going to look like until you get into the theater.
When we previewed it in December we found out that three (or more) drones made it feel a bit cramped and because of all the air they were kicking up, it started to interfere with their stability in flight.
We’ll also be redesigning the drones to be smaller, taking about 4-5 inches off the design is our goal. All the while we’re looking at ways to make them cost-effective.
And eventually we could look at ways to make the system rugged enough to go on tour and be operated by non-engineers.
We’re always looking for new ways to incorporate new technologies with artistic expression, so it’s possible that the Super Bowl performance gave our students some new ideas, we’ll see what happens and how drone performance evolves.
For more information about ExCITe’s collaboration with Parsons Dance and NextMove, visit: http://www.parsonsdance.org/
For media inquiries, contact Britt Faulstick, assistant director, media relations bef29@drexel.edu or 215-859-2617.
