A Special feature article written by Kashif Nawaz
IoT and Football
Internet-of-Things has created a buzzword for the past few months. Every website, article or news has some coverage of IoT and its’ application to a broader range of areas like healthcare, manufacturing, smart homes etc. In this article, we discuss the impact of IoT on the modern game of football.
- There is no doubt that modern day football has become an extremely complex sport. Gone are the days of players running in knee-length shorts and managers scribbling through cardboard slates. Not to mention that the training facilities, if now looked, could be considered medieval. If the reader just googles’ any “big” clubs’ training facilities (Man Utd, FC Bayern, FC Barcelona) he will he definitely appreciate how far we have come from playing Football in the mud and rain.
- This brings us to the main point of the use of technology in these games. Often, we have seen coaches/managers beaming onto their tablets scribbling, dragging positions, viewing complex formations, and analyzing complex cases, either pre-match or if required, during match time.
Big Data – More Data in Football
The movie Money-ball was iconic in the sense that it showed how data analysis, backed by continuous video performance and monitoring, turned out be a great game-changer for a small enough team to make them dream that they could go onto win the league and compete with the mighty Red Sox.
The principle – “each players’ performance can be quantified to usable, quantifiable metrics which are relatively absolute in terms of the assessment they have to offer and provide a way to both individually and relatively compare the players –
“It all boils down to a few numbers”
Moving from Basketball to Football and now almost, every other sport has embraced this concept. With hundreds of sensors, both on and off the ground, both in and near the players, it is possible for a coach to understand his players’ performance, their weaknesses, their lingering injuries and then optimize their performance or training routines.
This has also given rise to such a huge amount of data that even an ordinary fan, with just good access to the internet, is now able to break down the entire match into quantifiable numbers and statistics. It is not uncommon to hear phrases such as a number of passes allowed, total distance run in the same half, total distances covered in the opponents’ half etc. which were simply unheard of a few years ago. The reason: IoT.
With all these sensors tracking the players, not a single step is uncalculated or unaccounted for. Even in training, this is a great tool for coaches to understand the levels of players’ intensity, their tiredness levels or follow a player recuperating from an injury. Even heart rate sensors could potentially inform the medical training staff of a players’ vitals and alert them in case of any impending catastrophe to the player.
An IoT based Football Architecture
We will now review an IoT based football architecture based on the seminal work. For a more technical reading, the reader is advised to read the scientific paper.
(Ikram, Mohammed Abdulaziz et al. “Architecture of an IoT-based system for football supervision (IoT Football).” 2015 IEEE 2nd World Forum on the Internet of Things (WF-IoT)(2015): 69-74.)
We now describe each of the steps:
1.) Data Identification
With so many sensors attached to every player and so much of data available everywhere in the cloud, it becomes imperative to classify the data. By classification, we mean the first degree of classification, i.e. identification. How do we identify that the following available data comes from player X and not player Y? For this, we use something known as an RFID tag. RFID stands for Radio Frequency Identification.
Each sensor device is programmed to have a fixed RFID. This could be the players’ name or some internal code which could be used by the club to identify the player. In most of the cases, the RFID is simply a tag with a wireless interface that is capable of transmitting the players’ identification to the clubs’ servers and collecting the following data which comes associated with it.
2.) Data Collection (from the players’)
This is perhaps the most important step in the early stage of the system. Data collection is handled by the numerous sensors which are fixed onto the players’ body. These could be heart rate sensors, pressure reading sensors, accelerometers to measure the number of steps and distance travelled, temperature-based sensors, sensors on the payers’ heart, shin, boots, socks and knees.
Or these could also be integrated into a single (or two) sensors which are able to collect all the data. Each sensor is independent of each other and is responsible for collecting the data for which it is exclusively responsible. The most important aspect of this is continuity. The sensor must be able to collect data on a continuous basis to allow for complete and correct profiling during the measurement period.
3.) Data Collection (from the environment)
Simply collecting data from the players would not suffice. It is important to analyze the data from the surroundings such as weather and temperature conditions, barometric pressure levels, humidity and illuminations levels.
These allow the data collected from the players’ sensors to be integrated with the data coming from the ambient sensors to understand how the external weather conditions are affecting the performance of the players; how a certain player is responding to the given conditions and what changes need to be made in case of any issues.
4.) Sending data to a base tower
Most of the big clubs’ facilities are equipped with a huge server room of their own. In order to facilitate data streaming between the sensors and the servers, a base station (mobile tower) is available which collects the data from the incoming sensors and then sends it to the clubs’ servers.
This is mostly done as most of the sensors do not have the required intensity to send data to the servers directly (due to reasons such as the sensors themselves are pretty lightweight and comprise a few basic electronics).
The base station is able to aggregate the incoming data and then send it to the servers directly with much high intensity. Occasionally, the distance between a base station and a server may be a few kms whereas between the sensors and the base station could be a few hundreds of meters.
5.) Default Gateway
Occasionally, a single base station might not be able to collect all the data from all the players. Since the players could be moving around in different spatial locations at any instant of time, either in the stadium or the training facility, a greater number of base stations are required to collect all the incoming data.
These data are fed to a default gateway station which in itself is a one-stop shop to the main clubs’ servers. The GW stations directly connect the incoming data from the base station and send it to the servers.
6.) Cloud storage
The cloud (or simply the massive storage elements in the club’s facilities) are responsible for storing and processing all of the incoming data from the various gateways, base stations and sensors. The cloud storage has 2 main functions:
All of the incoming data is stored. If the club has huge facilities and is not short of cash, the data could be stored indefinitely to allow for a more thorough understanding of the clubs’ playing trends etc., the development of players over long periods of time etc.
However, most of the data is stored for a period of 5-10 years. Regular backups are mostly done so that in case of any issues, the data can be safely retrieved and restored.
The other main function of the Cloud is to process the data. With all the incoming data from various players and their sensors, the cloud is able to process the data coherently into organized formats such that they can be retrieved easily by the training and coaching staff based on the players’ names or any other attributes.
Complex algorithms run at the backend which allows for further processing and calculation from the incoming data. An example could be, how many players have their heart rate affected due to a sudden drop in temperature? All these kinds of queries which could be potentially useful to the staff are processed and handled by the cloud.
7.) Alarming Notifications
While the server is processing the data, it is also able to immediately calculate that any player could be suffering from injury or the extent of his/her injury is such that it could be fatal.
Of course, further medical judgment is required but the cloud is able to predict that a certain player might need rest or rehabilitation because a certain sensor (say from his/her knee) has reached a threshold value and using it beyond now could prove to the detrimental to the players’ health and physiological condition.
8.) Send Notification
Once the server has detected that a certain player is in physiological stress, it immediately sends a notification to the clubs’ medical and training staff and then they rush to the players’ care.
We have explained how a modern-day IoT based football club works. It is interesting to note that such a great value of effort using technology is invested by the clubs to take very good care of their players and hence, their performances.
IoT has enabled us to better understand, detect and prevent catastrophes which would on some other day, have resulted in very sad scenarios such as the death of the football Fabrice Muamba in 2012 due to a heart attack during a football match.
It is now left to the better judgment of the doctors’ and the coaches to understand and comprehend the data which is given to them by the system and utilize it both for enhancing the performance of the players as well as maintain their physiological well-being.
With IoT poised to become even more huge in the coming years (it is predicted that by 2020 we will have such 50 billion devices connected), it will also be possible for smaller clubs also to integrate IoT based systems in their curriculum and enhance their performance while maintaining the health and well-being of their players’ as a top priority.