Breaking News
Wednesday , August 23 2017
Home / BROADEN / Getting the Best Payback from IoT: Predictive Analytics

Getting the Best Payback from IoT: Predictive Analytics

The Internet of Things (IoT), which is a network of physical objects accessed through the Internet has big implications for both business and consumers. From data center environmental sensors to remote asset tracking, the IoT can help improve business processes and create new value. As technology continues to empower our ability to conduct analytics with “Big Data”, the “Internet of Things” has arisen as an area where devices themselves capture and transmit data albeit machine-level type data.

There are four fast paths to IoT payback: Connected Operations, Remote Operations, Predictive Analytics, and Predictive Maintenance. Analytics, from the simple to the complex, will play a big role in deriving value from IoT data. The sheer volume and potential velocity of data streams from sensors and other IoT devices will necessitate that analytics, such as predictive analytics, be used to both manage and analyze this data to generate payback.

First Path: Connect Devices, Sensors, and Meters, then to a Network.

Begin by connecting existing devices to your existing unified IP network and adjust your business process to take advantage of these newly connected things. When all the devices and facility are networked, you suddenly have a window into every part of the operation, enabling you to collect data from sensors that can help you to prevent part failures and optimize operations.

This is proven useful for manufacturers such as Rockwell Automation. By connecting assembly lines and operations in 20 manufacturing plants, and then connecting those plants to each other and its enterprise network, Rockwell was able to:

  1. Reduce inventory cycle from 120 days to 82 days
  2. Reduce rejected parts by 50%
  3. Increase on-time delivery from about 80% to 98%
  4. Avoid 30% in capital expenses

Second Path: Add Remote Operations, Monitoring, Control, and Asset Management.

Have you ever faced a situation where your packaging line suddenly stops and you get a safety alert that one of the many doors on the machine is open? You can’t restart the line until you send someone to the plant floor to physically check every door. However, implementing remote operations delivers a payback the first time it helps you avoid sending a person to see what’s going on.

In Nimble Wireless which is a startup in India has a client that operates 150 ice cream stores. The startup is often plagued by power outages that can ruin the inventory of a store in minutes and create a health hazard. To battle this problem, Nimble Wireless equipped the freezers in the stores with sensors that notify the manager if the temperature goes up 1 or 2 degrees.

The system keeps escalating the notifications all the way up the management line until the problem is fixed. It will even suggest actions to take, such as closing the freezer door or turning on the generator.

Third Path: Use Predictive Analytics to Identify, Understand, and Immediately Take Actions

Predictive analytics can help your staff sort and understand what’s coming in, so they can take intelligent actions. It can begin on a basic level, such as recognizing patterns and identifying exceptions can help you capture significant value. However, the real benefit of putting all these smart assets on the same IP network is to correlate and combine the data coming from multiple sources to gain new insights and to take even more valuable action.

Many industrial operations today run on a 24/7 basis. Downtime can costs plants a lot. By using the data coming from connected machines, combined with industry data and algorithms, predictive analytics can identify trends and probabilities of equipment failure within a certain timeframe. It also provides the opportunity to optimize the operations and take corrective actions.

Fourth Path: Adopt Predictive Maintenance to Increase Uptime

Things can break down at the worst times and in the most inconvenient places. If we can anticipate these breakdowns, the damage of the breakdowns can be minimized.

For example, Rio Tinto, a global mining company, operates extremely expensive equipment in very demanding and remote locations. The cost of just one of its enormous autonomous trucks being out of service is millions of dollars per day. That doubles if the breakdown occurs at the bottom of a pit mine and another vehicle has to haul it out. This cost can be minimized by predictive maintenance.

To minimize the cost of trucks being out of service, they implement sensors in each truck with 92 sensors monitoring engines, drivetrains, and wheels. The sensors generate data that is analyzed in near-real time to anticipate when the next failure is likely to occur. By this way, a truck can be rotated out of service for maintenance in an orderly way, instead of disrupting operations through an unplanned breakdown.

Solid and Proven Statistical and Probability Science

Predictive analytics is just solid, proven statistical and probability science. The longer the history and the more data you have, the more precise your predictions will be. From agriculture to health care to sports and entertainment, fast-payback IoT solutions can be adopted everywhere to get the best payback according to the different use case.


About editor

Leave a Reply

Your email address will not be published. Required fields are marked *