Smart Analytics is at the Heart of Successful IoT

0
240

Sensing and understanding the connected physical and digital world is the key challenge AGT as a company addresses since 2007. As a result of this work the IoTA analytics core library is the most advanced analytics framework for IoT. IoTA is built on 10 years’ worth of experience in sensing and understanding the “phygital” world. Our goal is to enable our clients, who operate across a wide range of sectors including connected cities, manufacturing, entertainment and sports, better understand their own specific environments so that they can answer key questions such as:

  1. Why can’t you catch a cab in Singapore at 3pm?
  2. When will China’s Yellow River flood?
  3. Am I producing enough solar energy for my house and still be able to sell some to my neighbor?
  4. Is the manufacturer’s machine going to fail in the next two weeks?
  5. Is this truck driver efficient or tired?
  6. Are the fans at the game chanting out of enthusiasm or are they potentially violent?
  7. Is this basketball player nervous?
  8. Is one bull more dangerous than another for this particular rider in an professional bull riding event?

To answer those questions, we have developed based on real world applications, a broad set of reusable IoT analytics modules that can be applied to a wide range of applications and data from different IoT devices. Our advanced analytics portfolio is designed to extract insights from both commercial IoT sensors (video, audio, acceleration, thermal and environmental) as well as consumer-oriented sensors like those in wearable items (smart bracelets, watches and shirts).

This article focuses on how the deployment of innovative (AGT) products, centered around smart analytics, can be used to create new user experiences/better outcomes in two select areas – (i) entertainment and sports and (ii) predictive maintenance for industrial machines.

In collaboration with WME/IMG (our Events partner with deep knowledge of/involvement in the entertainment and sports sector), AGT is currently sensing data from events, players, fans, referees and coaches in order to analyze/provide new insights on a broad array of areas that we/the event owner believes will be of significant interest to the “audience”, both live and extended. Examples include measuring/sensing the emotion of the fans, the intrinsic state of a coach, the behavioral patterns of an athlete in basketball, the force a rider in professional bull riding is experiencing as well as a measure for the level of aggression/intimidation that bulls in a PBR event are exhibiting.

Smart Analytics are built on top of advanced mechanisms, such as machine learning algorithms, to enable insights to be derived from the IoT sensor data based on the collection of large data sets used to train a model in a particular domain. For example, sound patterns of fans in stadium, are learned by a software model to derive behavior and finally emotions in real-time. These same principles can be applied to accelerometer measures from sensors mounted on bulls and riders to allow a score of a ride to be assembled and thus, to determine how tough the ride was and allow comparison between rides based on objective measures for areas that had previously only offered subjective results.

Those measures can be correlated with context data from the event, to create new content products interesting for the dedicated fan or the occasional viewer or even someone who has an interest in only an element of the sport, e.g. intimidation indexes. The content is consumed via apps, which serve as the communication front-end to the consumer. The merging of IoT data with pictures, videos and games allows the creation of a vast number of different/unique ways to view an event – the physical aspects of an event has been extended into a compelling digital experience.

In collaboration with partners in Industrie 4.0, we are applying advanced analytics to sensor data coming from machines such as molding machines or wind turbines. Relying on a huge set of historical sensor data, behavior analytics extract the normal and abnormal behavior of machines, and enables real-time prediction to determine if a certain part or a certain process is likely to fail. Those predictions can be used to avoid unplanned interruptions within the manufacturing process, and thus, reduce the cost and the risk for the machine owner.

Based on these same machine behavioral insights, and their typical failure patterns, the identification of design weaknesses can be discerned. This leads to constant improvement and potential innovations for the machine itself, resulting in a better, more reliable product for the respective brand.

The power of smart analytics in both of these domains demonstrate the potential that can be offered by collecting, analyzing and acting on the insights provided by IoT data. We contend that (almost) every application domain will benefit from advanced analytics built on IoT data sets. Platforms will be the foundation/home/launchpad for all these sorts of innovative applications. AGT’s platform, IoTA, currently far advanced in its production, is based on an infrastructure which scales from analytics at-the-edge (close to the sensor/machine) to large-scale implementations based on cloud technology; this dual approach broadens the ability to cover different scenarios/application requirements as well as speed of processing.

In the future world of increasingly blended physical and digital experiences, simply collecting lots of data and doing some statistical analysis will not cut it – information may be out-of-date by the time analyzed and remediation action is delayed and may not be relevant any longer. The winners will be those that act faster and with the knowledge that it is based on the right (time and specific case) relevant information. Thus, the need for leveraging ongoing real-time knowledge through actionable insights, based on real data and experiences, is why IoT/Big Data/Smart Analytics will be the key component/information necessity for all companies and organizations.