By Sharon Lin
With the advancement in technology, “internet of things” have become ubiquitous. This is especially true in the world of home automation. Home automation has been a popular feature in science fiction media for years. From robot vacuum in the Jetson’s to wearable computing in Back to the Future II, science fiction media has imagined what home automation will be like in the future. While some of these imaginations have not yet come to fruition, many home automation technologies have since become reality. Nowaday, companies try to automate everything in the house, from Nest’s self-learning smoke detectors and thermostats, to Dropcam’s motion-detecting home monitors. As more companies develop smart home devices, they are reshaping the way people live in their house.
Since the 1990’s, internet has become ever-present in many parts of the world. Hardware resource and development cost has significantly decreased. Along with that, the society is quickly becoming mobile-first. Because people nowaday are connected to the internet almost all the time via their mobile devices, this allows companies to collect vast amount of information about their users. It is this emergence of big data that really enabled companies to advance the development of home automation technologies. With a mobile device, data can be collected wherever the user is. Mobile device’s location tracking can document when a person gets home or when a person goes to work. Frequency of activity on mobile can also estimate when a person is on break or when the person is sleeping. Bluetooth technology can help track exactly where in the house the user is, whether that’s the kitchen or the living room. With such data, companies can develop home automation technologies that is programmable and self-learning. Aside from the dependence of mobile devices, many companies have incorporated sensors to collect data through the devices.
Combining data collected from mobile devices and hardware sensors allow companies to develop machine learning algorithms, which is the core of many home automation technologies. For example, with Nest’s Thermostat, it can learn a user’s living pattern – when does he leave for work and when does he come back home. From that, the thermostat can determine when to turn on and when to turn off. With continuous learning, the thermostat can adapt to different schedules a person have from day to day. Such device became a reality because of the vast amount of data the hardware and software collect in order to improve its predictive capability. In the case of Dropcam’s monitoring device, it uses motion detection algorithm to determine if there’s any unusual movement in the house when it should be empty. Through its self-learned schedule of the user along with the phone’s bluetooth capability, it is able to determine when it should turn off motion detection, so it is not alerting the user of unusual movement when the movement came from the user.
All of these are extremely important data to gather to further reinforce the self-learning within these products. However, privacy and reliability can be a major concern for people who are trying to adapt to home automation. For example, Lockitron is a product that allows user to unlocks a door using a mobile app. However, users may be concerned about how reliable it is for the software to be able to unlock for the user every time he arrives home. If the user’s mobile device is lost, they themselves will be locked out of their house, and they may be worried that someone will now have access to their house. Furthermore, in the situation where the mobile device is lost or there’s a security breach, the user’s data may be compromised. I believe these are all valid concerns. However, I also believe that big data is necessary for developers to keep on innovating. This is especially true when home automation products start to integrate with one another. For example, with a combination of home monitoring system and automatic locks, user can be notified if someone has entered their home, and they can remotely watch the activities within their home and notifies authority if necessary. However, it can also be dangerous when companies start sharing user’s data with one another. There should be regulations regarding what kind of data is allowed to be shared, and what kind of companies should be able to receive such data. This gives some protection to user’s data, while allowing companies to keep innovating.