When the Referee Isn’t The Only Zebra On the Field
From the very naming of its business, it would appear the Lincolnshire, Illinois based firm Zebra Technologies has a way of visualizing the product it delivers. Specifically, when it was founded in 1969, Zebra focused on implementing barcode and UPC technology for industrial use. Specifically, in creating these barcode tags, it provided businesses the ability to track whatever might be critical to its internal operations: inventory management, package tracking, etc. The visualization of this, of course, played out in the very name of the company, as the barcodes themselves represented zebra stripes.
In 2007, however, Zebra acquired another business – WhereNet – which specialized in making a much more high tech version of the barcodes Zebra had built its business upon. These Radio Frequency Identification (RFID) tags allowed for a much more dynamic type of tracking to occur, as unique RF pulses could be emitted from different tags, providing real time updates on locational data on items the tags were attached to; the potential for this digitization of a previously analog process promised to fundamentally change the trajectory of the business. After Zebra’s acquisition of WhereNet, a string of further investment in locational software and hardware drove the company to think about new potential ahem – arenas – it might be able to compete in. The new target? As Jill Stein, VP and GM of Zebra’s Location Solutions puts it: “We started a project in sports. 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.”
Beyond the ability of the RFID tags to rapidly transmit locational data, however, the data would be significantly less valuable without the various other investments Zebra made in the wake of acquiring WhereNet. Zebra is able to truly capitalize on the value of these rapid emissions due to the low latency – or delay from signal to server – as it only takes 120 milliseconds for RF emissions to reach the main server after departure from the tag. The result? Locational data accurate to an astounding 6 inches wherever a tag is located!
Having made the appropriate investments to track at the speed and granularity required to provide meaningful data in an athletic setting, Zebra approached the National Football League about partnering in a way that might bring the benefits of this digitization to NFL players and fans alike. The partnership began to take hold during the 2013 season, when over 2,000 players were outfitted with tags, and 18 of the NFL’s 31 stadiums installed the requisite receivers to track the RFID emissions.
This relationship continued to blossom into the 2014 NFL season, as the Zebra real time tracking system proved its ability to harness digitization to the league’s benefit. Over 2,000 players once again received sensors, with one quarter-sized RFID tag placed under the left and right shoulder pads of every player. (As a reference point for those who have never put on shoulder pads – pads are constructed such that the tag neither directly touches the player’s body nor is placed on the outside of the pads in a way that would affect the contour or outline of the player). By having two tags on each player, directional data could be deployed – specifically enabling the sensors to shed light more precisely on how and when a player turned. The net output of all these sensors for the 2014 season was a whopping 1,692,000,000 distinct XY player coordinates.
While the sort of insights afforded by Zebra’s digitization of on-field player tracking are in the phases of gaining traction for the NFL’s fan base (e.g. at-home fans are now able to see in either post-play or post-game analysis the exact routes run by players on the field), the growth in this area holds a great deal of promise. Employing applications through Windows 10 and XBOX One, Zebra has sought ways to tie highlight clips to individual players’ data (specifically, through an app called “Next Gen Replay”) and, in an even earlier implementation of the technology, Zebra was able to provide in real time, on the stadium’s display, which players were on the field at any given moment.
For all the benefits fans might reap from this technology, however, NFL teams and players have stood to gain even more. Through a more detailed analysis of how players are precisely moving on the field, these athletes are able to not only focus on their individual performance, but teams can analyze their collective performance with a more granular set of data that simply cannot be derived from game film.
Note: all of the information and graphic for this blog entry drawn from “The Internet of Things Comes to the NFL” [http://www.cio.com/article/2980853/wearable-technology/the-internet-of-things-comes-to-the-nfl.html]
Words: 788
This is a great application of the technology. Does the placement of the RFIDs impact the type of data collected? With football being such a high impact sport, how do they ensure the RFIDs stay on during the game?
The additional element that makes this whole system go is Zebra’s ability to get an RFID chip small enough to fit inside of a standard NFL football without altering the flight or rotation of the ball. This is essential to making this system work from a scouting and player evaluation standpoint. If teams don’t have a clue where the ball was on the field for that particular play, then the data is fairly useless beyond a sports science and health application. The other element that is a big issue that the NFL is still trying to work through is data management given this influx of information that needs to be stored and protected. There is currently a wide breadth of technical aptitude within team front offices so the NFL is taking a very cautious approach to releasing this data to ensure that all teams are at least capable of managing this new information before it is given to all clubs. Really great article and if you are interested in learning more the NBA has been utilizing this concept of player tracking for a while.
http://www.nba.com/2016/news/09/22/nba-stats-deal-sportradar-second-spectrum.ap/
Very interesting discussion on how tracking technology is being applied in the NFL. I question what types of insights analysis of the data will allow. Apparently teams do not yet have access to all of the data that the technology is collecting for purposes of improvement in playing; the NFL is currently analyzing the data for actionable insights so as to provide teams with solid data-supported recommendations (https://www.wired.com/2016/01/the-nfls-impending-data-revolution/). As the article mentions, providing all of the raw data to teams may be confusing and it may be unclear how it can be used. Thus, after statistical analyses have been performed, it would be interesting to understand whether the NFL is working on building a software platform with easy-to-use GUIs that would make the data and data analysis more accessible to players and teams. Similar to most applications of big data, collecting it is the easy part; understanding how to use it is more challenging.
This is a very intriguing topic to me. Between this technology and what the NBA and FIFA is able to do with the SportVU cameras, I am consistently amazed by how technology is changing how we are able to analyze what we see on the field.
Part of my interest into this topic comes from the idea that it could impact the worth of coaches and the amount of value they have on the field. Theoretically, as more teams gain the ability to intelligently analyze what their players are doing on the field, it could level the playing field for coaching and lower the barriers to entry for the profession. It will also likely result in a shift from coaches that make their decisions based on “feel for the game” to those that have the ability to analyze the data they receive and make strategic adjustments based on what the numbers tell them.
On a completely different note, this article also made me a bit concerned regarding the application of this technology in other areas. If the cost of inserting this type of technology is fairly low and it is something that can be completed fairly easily at scale, one could feasibly start to put sensors in the clothes of normal people. While I assume there are a number of privacy laws that would prevent this from happening, even the prospect is enough to spark a little concern from me.