Scroll to top
© 2018, Titus Kipruto

How to Change your Data Collection Approach for IoT

admin - August 17, 2017 - 0 comments

Tech industry has realized that data will play vital role in the business strategy. Since the tendency is towards digital transformation, the importance of data can only increase in future. Internet of Things(IoT) which thrives on real-time data is the parallel trend in this disruption. When dependency on IoT is about to witness the utmost scope in urban lifestyle, it is important to optimize your data collection methods to offer better results.

Primary misconception we get into when we undertake data collection in IoT is holding all the data with the same importance. Ideally it is not so simple. Some data set demands higher degree of diligence. When we seek the data from a source, we often don’t think of its actual utility.

But if we can reverse engineer and figure out which is the best data, we would abstract that the data which gives you actionable insights is the one which is best. The data which don’t help you to interpret and prompts you take the further action is just a junk or obsolete in many cases.

What mistake many companies does in Data Collection?

When we want to collect data, we easily jump into a conclusion that more and more we collect the data, better the chances are there to excel in the business. In that verge, we tend to solicit all kinds of data from devices which can be pulled in. Hence this poses to be a quantitative approach rather than being quality based.

A superior way of handling this is by first putting up a question “What problem my business is going to solve to my target clients?”. This will enlighten you on what kind of data you should seek and eliminates the possibility of data getting siloed. More the data you have is ambiguous, lesser the actionable insights they can offer you.

Consider a case where you have to draw the periodic data of the blood pressure patient who is wearing a tracker around his wrist. You don’t need to pull the data every second. If not the data set will become enormous that will become in itself a roadblock effective predictive analysis. Conventionally the data in this case should be drawn once or twice per day to know how the patient is maintaining his lifestyle.

But, you cannot always expect that the patient follows same lifestyle throughout the full year. He will have many variations like hanging out in pubs and playing games etc. In that case convention of drawing data twice a day will not suffice. That will not prove the purpose of wearable IoT either. Hence it is then good to resort to Edge Analytics to know the spikes in the user behavior. This can be followed in any business which is going to depend on IoT.

Going forward, the data will exponentially surge to the unprecedented amount. Hence there is nothing great in having huge bulk of data which will not help. Current era is not of having everything, but selecting what actually you want and how you can interpret it and innovate using it. It will be the case IoT data collection too i.e if you know how to translate data into actionable insights, follow the thumb rule ‘Less is More’

Post a Comment

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