Making Money from Data Analytics of a Connected Home.


Staff member
Apr 23, 2020
Received Likes
Making money from data analytics of a connected home.

IoT is a field where a person can make money by providing services or by providing products which make the end users life easier. In this guide, we will be discussing about data analytics from the data of sensors from a connected house.

In today’s world we have devices like Google Nest, Alexa etc. which are capable of controlling lights and appliances with just a voice command. But these devices aren’t capable of taking these decisions on their own and performing the actions automatically. For such tasks we need a connected home.

A connected home is networked to enable the interconnection and interoperability of multiple devices, services and apps, ranging from communications and entertainment to healthcare, security and home automation. These services and apps are delivered over multiple interlinked and integrated devices, sensors, tools and platforms. Connected, real-time, smart and contextual experiences are provided for the household inhabitants, and individuals are enabled to control and monitor the home remotely as well as within it. Providing such experiences requires a lot of data analytics, and the data produced by the sensors and devices is huge which helps in developing many important experiences for making users life much easier.


Goal is to gather data from a connected home and providing the service of determining the user behaviour and activities and enabling smart decisions on their houses according to the users activities. After analysing the data, the model should automatically take decisions of user activities like switching lights and AC ON before the users arrival, and setting home AC temperature same as that of users car, etc.
The target audience of this idea will be nearby neighbouring households who are interested in making their lives easier and businesses who are ready to pay for the data of connected home sensors or for the analytics of the usage behaviour of the user.

Step 1: Identifying Demands

Nearby societies and residences where most millennials and tech savvy people reside. Also the families having family members of the age group 0-5 years or 70-90 years are the ones who care most about the family. Such families and residences can be approached for providing the service.

Companies who are working on making smart homes and connected house, car, etc. need a huge amount of data of user behavioural activities for enabling their products to take smart decisions.

Step 2: Gathering Data

A pre existing simple smart house is needed and a hub needs to be attached to the existing system. The hub needs to be configured such that it sends data to you. Data transfer can be done in different ways, the best way could be to send and store data on some inexpensive cloud platforms like ThingsBoard instead of going with AWS or GCP. The data can be accessed anytime and can be used for further processes.

The data needs to be gathered for a period of more than 2 months so that the analytic can be more effective and precise as the analytic needs to predict user behaviour on hour basis.

Step 3: Data Pre Processing

Raw data available in the cloud needs to be improvised. Many times sensors fail to capture data and return null values, and you need to refine those null values. It wouldn’t be appropriate to drop an entire column for just a few percentage of null values present in the data. A good practice is to replace those null values with mean or median value of the entire column. Here these columns have data from the sensors placed at different locations of the house.

Step 4: Predictive Analysis

The time series model comprises a sequence of data points captured, using time as the
input parameter. It uses the last year of data to develop a numerical metric and predicts the next three to six weeks of data using that metric. It is a potent means of understanding the way a singular metric is developing over time with a level of accuracy beyond simple averages. It also takes into account seasons of the year or events that could impact the metric.
Predictive analysis can be used for predicting the user’s general activity at that moment of time. Also, many other machine learning and deep learning models are capable of predicting this. It’s up to you to decide the best algorithm to use.

Step 5: Enabling actions according to the predictions

Once you have achieved a good rate of precision in predicting the user’s activities then you can decide to help the user with his activities. Like making TV automatically at the time of his favourite show, enabling lights and AC before the time of his reaching home, etc. For enabling these actions you need to send these as commands to the hub which you installed in the house which controls the sensors and devices. The hub will receive commands from the cloud and will enact on them accordingly.

Step 6: Selling the service

You can try to sell this service as a subscription based model to the residents of societies and to millennials. You can also sell the data and analytics to businesses on a subscription based model, or a product based model for a specific amount of data. You can also charge for the cloud services in your service charges.

Hope this guide provided a way to earn money just by making a few simple things and some data analysis.
Last edited by a moderator: