Arcadia Data is honored to have been selected as a “cool vendor” in IoT analytics. This connected device revolution is driving innovation and efficiencies in our global economy.
Check out what Gartner has to say about Arcadia Data in their research report:
Cool Vendors in Internet of Things Analytics, 2017*
A Gartner login is required to view the report.
The IoT (internet of things) has been greatly hyped by the media. Many of us think about “personal IoT” devices such as FitBit, which provide us analytics on the number of steps we take, or whether we had restful sleep. But what about the broader usage of connected devices in industrial IoT applications such as predictive maintenance, anomaly detection in the manufacturing process, or analyzing product usage patterns to drive product development? Gartner estimates more than $440 billion will be spent on IoT in 2020 across all industries.
So, where does a business start, and how do they show value from IoT applications – to separate hype from business value? Analytics, of course.
Garter released their Cool Vendors in IoT Analytics report last week which discusses the importance of analytics providing visibility into the impact of the IoT on the business and empowering individuals. In the report Gartner states,
“Analytics enables the most visible impact of the IoT. It drives new business scenarios and empowers individuals. The possibilities for analytics are limitless, and ever-growing data sources present more opportunities to innovate through holding novel insights. But currently, data and analytics leaders should narrow their focus to deriving additional value from the ongoing IoT implementations. Examples are an increase of the equipment lifespan, asset optimization, predictive maintenance of devices, anomaly detection in the manufacturing processes or finding new product usage patterns.”*
But we’ve had business analytics for many decades now. So how is the IoT different?
In many use cases, customers and practitioners are finding that the scale, velocity, and complexity of data in IoT applications breaks existing data analytics systems. Gartner states:
“In many cases, general-purpose analytics tools can be applied to the IoT, but there are two differences specific to IoT analytics:
- Massive amounts of sensor data, which is coming at high speed, mostly in time series. This data often must be enriched by other data from various sources for more-accurate analysis (for example, warranties, claims and domain ontologies combined with sensor data from the equipment that provide the information required for precise breakdown diagnostics).
- New analytics users — mainly citizen analysts and engineers — who have an in-depth domain expertise that will allow them to interpret and incorporate insights from IoT implementations. Citizen analysts are IoT project members who want to easily visualize operations in their domains. Engineers working on IoT projects prefer to code. They look for the means of embedding analytics into their products and processes to make them smarter.”
From the perspective of Arcadia Data customers, we see five key tenets to providing valuable insights from the IoT which are different from business as usual.
- Access to Real Time – take an example of remote product monitoring of hardware servers at a customer site. As a manufacturer, you’re interested in analyzing the historical longevity of your product – what parts are failing, how often, and are there indications that can predict when a part will fail? But being able to apply this historical trending analysis in real-time for customers is where real value is provided to customers so the manufacturer can provide proactive alerting before something fails and take remediation actions. IoT analytics need to be able to monitor, analyze, and alert on top of machine-generated logs quickly — without complex data pipelines and data preparation.
- Combining both operational real time “silos” with historical “silos” in one “pane of glass” – IoT applications open more real-time monitoring scenarios such as in connected vehicles where you may want to take immediate action on events as they happen such as excessive speeding, long idle times, or an accident, combine that with historical information and ultimately decide whether or not to insure the vehicle based on specific driver habits and risk profiles. In this example, fleet managers would use a legacy tool or homegrown tool for monitoring activity in real time, but would use traditional BI and analytics tools to create dashboards and reports to measure key historical metrics of vehicles, drivers, and environmental factors. This slows the time to action by requiring “swivel chair” analytics between tools even when the data being used for both these analyses is the same. Juxtaposing real-time and historical information into IoT applications such as this connected vehicle demonstration video give operators a fast and efficient way to manage their fleet.
- Large scale of data analysis – sensor devices on manufacturing equipment and transportation alone are causing an explosion of data volume. The new Boeing 787 generates half a terabyte of data per flight! Imagine the data storage for a fleet taking hundreds or thousands per day. Storing, processing, and analyzing this data cost-effectively for predictive maintenance or other applications requires a new approach over traditional data analytics architectures. Our customers want to store AND analyze the data in one place without the additional cost of extracting data down to a size manageable by traditional systems.
- Flexible deployment in cloud and on premises – connected devices by definition will be in multiple places and require data movement and synchronization across the network of “things”. The data, therefore, is likely to be in the public clouds as well as centralized regionally or centrally into private data centers. Technology platforms that work seamlessly across cloud and on-premise deployment while still maintaining the same application interface are important. In addition, visual analytic tools which have network analytics to show interconnections across devices are handy in this use case.
- Support embedded analytics within applications – Lastly, IoT data is valuable when embedded into applications. Sometimes, analytics IS the application in the IoT. For example, consider Uber’s new data service, Uber Movement which provides anonymized data from over two billion trips to help urban planners around the world. With Arcadia Data, embedded analytics is easy to do through programmatic access to our visual platform.
Supporting these key criteria for IoT analytics provide visibility across the entire connected web of devices to drive business value to customers, suppliers, and partners. A great example of this is a Fortune 500 customer using Arcadia Enterprise for data center monitoring who was able to provide secure data access all the way down to individual logs across many different users including product managers (what features do customers use most?), customer support (what are the logs associated with a particular device for a service call?), as well as sales and marketing to better understand their customers’ usage of products in the field.