The Internet of Things comes to the NFL
On Thursday, when the defending Superbowl XLIX champion New England Patriots host the Pittsburgh Steelers to open the 2015 football season, each player will be equipped with a set of RFID sensors about the size of a quarter embedded in his shoulder pads, each emitting unique radio frequencies. Gillette Stadium (and every other stadium used by the NFL) has been equipped with 20 receivers to pick up those radio frequencies and pinpoint every player's field position, speed, distance traveled and acceleration in real time.
By using two sensors for each player — one embedded in the left shoulder pad and one on the right — the system will also be able to identify the facing of each player.
[ Related: NFL teams test virtual reality at 2015 training camp ]
The NFL plans to use the data generated to power the NFL 2015 app for Xbox One and Windows 10, allowing for things like "Next Gen Replay" that will allow fans to call up stats for each player tied into highlight clips posted on the app. But that's just the beginning. The data will be fed to broadcasters, leveraged for in-stadium displays and provided to coaching staff and players.
"We've always had these traditional NFL stats," says Matt Swensson, senior director of Emerging Products and Technology at the NFL. "The league has been very interested in trying to broaden that and bring new statistics to the fans. Along the way, there's been more realization about how the data can be leveraged to make workflow more efficient around the game."
"This type of initiative really opens the doors to do more things at the venue," Swensson adds. "At the Pro Bowl last year, we had a display up that showed what players were on the field. By putting up what players were on the field in real time, it really gave fans more information."
Lincolnshire, Ill.-based Zebra Technologies is the NFL's technology partner in its IoT push. Founded in 1969, Zebra built its name on manufacturing and selling marking, tracking and printing technologies ranging from thermal barcode label and receipt printers, RFID smart label printer/encoders and card and kiosk printers. Beginning in 2013, it began pivoting into IoT and machine to machine (M2M) applications with the launch of its Zatar software platform. In the same year, it launched its MotionWorks Sports Solution, which powers the NFL initiative.
[ Related: Fan experience key to success of drones, VR in pro sports ]
"Traditionally, Zebra has done bar coding," says Jill Stelfox, vice president and general manager, Location Solutions, Zebra Technologies.
But in 2007, she explains, Zebra acquired WhereNet, a specialist in active RFID tags. A number of acquisitions in real-time location software and hardware followed, and in 2012 the company began thinking about new markets that might be accessible to it through the technology.
"We started a project in sports," she says. "What do you need in order to effectively track professional athletes You need the ability to track a motion in subseconds. Our tags can blink up to 85 times per second."
You also need the capability to deliver data from a tag to a server with very low latency. She notes that it takes about 120 milliseconds between the time a tag blinks on the field and when it hits a server. The location data is accurate to within six inches.
"Then you need to put analytics on it so you can make decisions about what's happening in real time in less than a second," she adds.
With a hardware and software solution in hand, Stelfox says Zebra had to find a customer. It decided to go after the biggest fish in the pond: the National Football League. It started with a meeting in December 2012, before the Zebra Sports Solutions unit had even launched. The partners went live with a trial in 2013, equipping more than 2,000 players with the tags and 18 of the league's 31 stadiums with receivers. Tags were also deployed on officials, yard markers and pylons.
[ Related: IBM Watson brings cognitive computing to fantasy football ]
Over the course of the season, more than 17,000 plays of NextGen Stats were measured — more than 1.7 billion sets of XY player coordinates measured, transmitted and stored during games.
"Every NFL stadium is connected to a command center here in San Jose," Stelfox says. "That command center has to operate as sort of a central command of all the data. When the data is collected in the stadium, it's sent in the stadium to the broadcaster in the stadium — it never leaves the stadium from a broadcaster perspective — but it's also distributed out to the NFL cloud."
All that happens in under a couple of seconds.
"The command center is our point of clarity," she says. "We can see every tag on every player from San Jose when the game is live. If there's something that goes wrong, we know about it very quickly and we have dual recovery. All of that is controlled from a single point of coverage in San Jose."
For coaching staff and players, the technology won't yet be available during games.
"Initially, it's really more of the post-game," Swensson says. "Right now, we have a lot of stuff going on on the sidelines. It could just be too much of a distraction during the game. It might be a place we get down the line, but right now it's not what we're trying to solve for."
That said, the data will be available to coaching staff and players who could use it for post-game evaluations. It may also play a role in training.
"We've just scratched the surface of what we can do with the data," Swensson says. "Every week there's another thought about how we can expand upon the information we've pulled together."
"The possibilities are truly endless," adds Stelfox. "The players love this kind of tracking technology because of that. They're professional athletes by every stretch of the imagination. They want more data about themselves — how they can stay hydrated better, perform better. Anything that can help them do that, they really want."
She notes that the New Orleans Saints and Head Coach Sean Payton have worked closely with Zebra on the player performance side of things.
"When you tell him that his star running back just ran seven miles in what was supposed to be their off-day practice, that's probably four miles too many," Stelfox says. "He might come to us then and say, 'I want you to tell me when he's gotten to four."
One of the big lessons Zebra learned in the process, she adds, is that generic data scientists weren't sufficient to gain insights from the data. Zebra needed football experts.
"When you look at analytics in football, you really need football people," she says. "We had to go out and hire football people. The analytics from manufacturing weren't the same as the analytics from football. We could see correlations in the data that seemed important and then they weren't. We had to bring in people that had the football expertise who could say, 'Look, this is why it matters.'"
This season, Swensson says, you can expect the NFL to surface some of its new data on its Thursday night broadcasts and portions of games aired on CBS. The data will also be integrated with NFL.com and the NFL's fantasy football offerings.