Internet of Things (IoT) and Real-Time Analytics - A Marriage Made in Heaven
The Internet of Things provides a constant stream of data, making real-time analytics the perfect tool to analyze it.
The Internet of Things (IoT) represents a creative disruption, something that begins to topple existing processes and technologies and brings forth a completely new way of working. IoT can usher in improved products and services, customer experience, security and healthcare, among other things, if it is properly harnessed. One of the best ways to harness its full power is real-time analytics. IoT and real-time analytics constitute a package. Without real-time analytics, you cannot harness the full benefits IoT has to offer. IoT complements real-time analytics and vice versa. However, to combine IoT and real-time analytics, organizations need to make a lot of changes in the way they currently go about business.
|Webinar: Standing at the Edge: Streaming Analytics in Action
IoT and Real-Time Analytics Use Case
The driverless car seems to be an appropriate use case for the combination of real-time analytics and IoT. A driverless car is fitted with several sensors and an IP address. When a driverless car travels down the road, how does it interact with other things on the road such as traffic signals and other vehicles? The driverless car will generate and relay data as it travels; this data includes information such as speed, time to reach certain landmarks and emission percentage. Given below are some possible influences over driverless cars:
- The driverless car will receive analytics from traffic signal points on the traffic congestion in the city. Based on these reports, the car can automatically choose the route with the least congestion.
- The nearest traffic signal points will send data on the time remaining before the signal turns red. Based on the data, the driverless car can adjust its speed.
- Traffic police can receive reports if the car is traveling above the permissible speed limits. This will trigger a notification and the car will be stopped at the next control point.
- The pollution control authority of the city will receive the emission data and send a notification to the owner of the car if the emission percentage is above acceptable limits.
- As the driverless car reaches its destination and searches for a parking space, its sensors can quickly scan and find vacant spaces, if any.
So, what are the findings from the above use case?
- To make sense of the data generated by the car, it needs to be received in real time.
- There need to be several other sensors, such as those in the traffic signals and pollution control offices that receive the data in real time, process it, create analytics out of it and trigger an action such as sending a high-emission-level warning.
- Without real-time analytics infrastructure, receiving IoT data does not make any sense.
Industry Attitude Toward IoT and Real-Time Analytics
It seems that the industry has been embracing the powerful combination of IoT and real-time analytics, and there is a lot of optimism surrounding it. In a survey conducted by Vitria, an advanced analytics solutions provider, it was found that 48% of the respondents had already been working on IoT and real-time analytics projects. The respondents replied that they were actively investing in IoT and real-time analytics. Two things emerged from the survey:
- Real-time analysis of the data generated by IoT devices were of prime importance.
- Companies are depending a lot on the predictive insights given by real-time analytics.
The salient findings from the survey are:
- Mobile devices (32 percent), smart meters, cell towers and sensors fitted in vehicles and logistics points are biggest sources of IoT data.
- 48 percent of the respondents are working on active projects while 15 percent of the respondents said that they have worked on it within the past year.
- 43 percent of the respondents said that they would invest in IoT analytics, automation and visualization, while for each area separately, the response was IoT analytics (20 percent), automation (8 percent) and visualization (5 percent).
- Business intelligence is the area where streaming analytics is being used the most.
- 18 percent of the respondents said that they paid the highest priority to predictive maintenance, while 17 percent said that they needed real-time analytics for network monitoring and service assurance. Only 8 percent said that they needed the solution for field service management.
- Most investors foresee IoT and real-time analytics providing a lot of value in the future.
Returns on Investment on Real-Time Analytics and IoT
The paragraph above seems to paint a rosy picture of the real-time analytics and IoT team. Many experts are talking as if the combination is a panacea. The answer is not so straightforward. The industry needs to see past the hype and realize that a lot of hard work is in order to get significant returns out of the real-time analytics and IoT combination. That does not mean that the combination is a bubble, about to burst; there is a lot of substance, it is just that a lot of work is needed. Let's take a look at what we need to do in order to maximize the returns. Let's think about the primary steps:
Identify the Reasons You Need Real-Time Analytics and IoT
Identify the business problems you want the real-time analytics and IoT combination to solve. For example, a company working on environmental protection and conservation may need real-time analytics on the amount of carbon dioxide emissions in a city during peak and off-peak traffic hours. It may also need real-time analytics on the number of vehicles exceeding the permissible emission limits.
After you have identified the problems, decide if the real-time analytics and IoT combination is the best available solution. Do not be influenced by the hype — base your decision on cold facts.
Estimate the Costs
After you have identified the problems, conduct an objective, data-based ROI analysis. You should, among other things, focus on two things: the total cost of ownership and the benefits you are likely to derive. The key to a successful analysis is having quantitative outputs from the analysis, as much as possible. For example, the IoT and real-time analytics should be able to predict the time frame in which machinery in your factory will start giving diminishing returns. This is also known as predictive maintenance. Secondly, find the total cost of ownership which includes, but may not be limited to, the people you employ for this assignment, equipment such as computers and servers, training cost and time and maintenance of sensors.
Understand the Challenges
Implementing a real-time analytics and IoT project is a huge and and extremely complex undertaking because for most organizations, it is unprecedented. It is important to do a realistic assessment of the tasks and break them into smaller, manageable chunks.
The first step toward getting the best out of the combination of real-time analytics and IoT is to accept that it is no magic wand. At the same time, it is not a bubble. Avoid extreme thoughts. There is a lot of substance in the concept, which needs to be harnessed carefully. You need a realistic assessment and quantitative analysis followed by small steps. This is a project which could redefine your business like never before if you can implement it properly, but it is going to take time.