International Data Corporation (IDC) predicts that the worldwide market for IoT solutions will grow from $1.9 trillion in 2013 to $7.1 trillion in 2020. This forecast corresponds to a predicted number of connected devices of between 40 billion (ABI Research) and 50 billion (Cisco). As more and more devices begin autonomously sending and receiving data, the sheer volume of data in the world is forecast to grow ten-fold by 2020 (IDC), from about 4.4 zettabytes to 44 zettabytes. The amount of information produced solely by IoT devices will account for about 10% of all data. Managing the complexities of extreme data volume and velocity is paramount.
There is really no difference in the fundamentals of real-time stream analytics when it comes to the Internet of Things. However, IoT exacerbates the challenges of real-time stream analytics due to the sheer data volume and the increased variety of data sources.
The dramatic increase in data sources leads to greater variance in data formats (structured, unstructured, varied hierarchies, etc.) with little standardization. Solutions must correlate among numerous streaming and static data sources, and involve contextual information in the analytics. And IoT data has a higher likelihood of being time-sensitive, therefore requiring the ability to process the data sequentially.
The ultimate risk to IoT companies of being unable to process, visualize and act upon such high-volume, high-velocity data is simply being “Too Late”. Too late to prevent a shipment of fresh produce from spoiling. Too late to understand that half of the sensors in an oil field are offline. Too late to alert someone that their car has an oil leak before it throws a rod. Too late to make a marketing offer at an emotional inflection point. Too late to correlate multiple events to realize there’s been a security breach on a wearable device.
The robust nature of the Striim platform and its ability to assimilate, process, analyze and take action on the highest volume and highest velocity data yields many advantages for IoT providers. The platform can assimilate both structured and unstructured high data volume from a wide variety of real-time IoT sources such as IoT and mobile devices, sensors, log files, and geolocation services. Additional data sources can be added on-the-fly.
The massive flow of device and application data in an IoT environment makes it absolutely necessary to make sense of this data instantly as it streams through the system. There is no time to offload data for analysis. Our platform enables IoT vendors to process data both sequentially and continuously as it is received, aggregating, correlating and analyzing streaming events as they happen to provide meaningful visibility and automated response.
For example, IoT devices are often used for fleet management in the shipping and logistics industry. Through such devices, central dispatch can monitor the exact location of a truck at any given moment, as well as assess the speed at which it is traveling, fuel level, tire pressure, wind speed, etc. to determine a highly accurate estimated arrival time.
Critical to IoT analytics, the platform is able to correlate streaming data with both historical and contextual data. So, for example, a set of device readings is continuously compared to a baseline, and at one point are found to be outside a predefined bandwidth. Contextual data such as temperature or humidity readings shed further light on the situation, indicating the devices are too warm to take accurate readings. The situation is automatically resolved as the system triggers an air conditioner in that zone of sensors.
The volume and velocity of IoT data today and in the future increases the complexity of correlating and analyzing this information across a wide variety of devices, data sources, data formats and environments. Striim handles these complexities, enabling companies to perform the multi-stream correlations and analytics they require, with all of the speed and context of reality. These companies are dramatically increased the usefulness of their data, leading to more efficient operations, happier customers, and greater success of their IoT offerings.
Operational efficiencies are mostly centered on the ability to address issues before they affect the business. For example, understanding early-on that the power usage of a connected car device is getting close to draining the car battery can mean the difference between placing the device in power-save mode, and having to send roadside assistance out to jump-start the stranded car.
A happy IoT customer is one who constantly experiences the expected level of value and service from the IoT device. This can only be achieved through the continuous monitoring of the device and service, early detection of anomalies, and automated response to issues.
In general, the overall success of an IoT product hinges on the vendor’s ability to provide the service in an optimized manner, and address issues at the speed and in the context of the real-time customer experience. Striim makes it possible for IoT vendors to provide a “Never Too Late” culture and offering.