appiOT flow

Internet of Things – ioT

Machines, cars, smart appliances, wearable health monitoring devices, and many others have one thing in common – They all have a small computer with networking capability. That means they can send out over the internet the data they generate. Internet of Things is about collecting data from things of interest, store and put the data to use. Putting the data to statistical analysis can yield useful information for decision making.

If a car sends out its odometer reading, any receiving computer system can suggest oil change every 3000 miles and transmission oil change every 30,000 miles. These are deterministic. It would be useful if we get info such as “If you do not change your catalytic converter now, you’ll end up in a stalled engine in 5 months”. Statistical analysis would be needed to analyze trends in data collected over a long period of time, or large volumes of data collected from multiple “Things” collected over a short period of time. Machine Learning based algorithms, which get better over time, with more historical data, would answer futuristic questions.

Some of the things Enterprises want to do are – Suggest complementing products to on-line consumers during check out, send out discount coupons to the right set of customers for the products they are interested in, compute “possible breakdown” of machines and inform engineers in advance, judge a job applicant on their propensity to stay more than 5 years with the organization, forecast demand for their products, etc

appIOT logo

appiOT

Many analytic techniques are available. The crux is to find the right analysis method for the objective, for which we set out to collect data. Sometimes we might find a best fit line, curve or graph to best describe the spread of data collected. Other times we may resort to “Machine Learning” techniques or algorithms to answer futuristic questions or group (classify, cluster) the data.

Practical Use - Schedule an unscheduled Service Call

A machineries manufacturer schedules periodic maintenance visits by service personnel to Customer Locations. With appIOT, the same manufacturer, collects measurements from its install base machines into the data lake, analyze for “Service Needed” indicators (predictions), and schedules preventive maintenance visit by service personnel in its Operational ERP System.