How an AI like Jarvis will redefine IoT
Technology evolution will reach a new height with AI as Facebook steps in with its own IoT product: Jarvis. No readers!! We are not talking about the beloved Jarvis from the movie series Iron Man; we are here discussing the new AI product from Facebook which is set to redefine the world IoT and will take it to the next level.
Jarvis is built with a sole motive of building an AI powered device effective to run home appliances just like Jarvis does in the movie Iron Man. FB has made a strong point to go beyond Siri and Echo and make IoT device more than a ‘Do as I say machine’.
Though Mark Zuckerberg made a New Year resolution to make simple AI run home devices, the project turned out to be more difficult. He struggled in finding the apt devices whose power can be remotely controlled by an app and had to find out from the nuances of apps to figure out the things that would work perfectly with the IoT device.
Before Jarvis, Mark was working on connecting his phone and computer through AI that can control the home including lights, temperature inside, music and security. With Jarvis they use several Artificial Intelligence techniques like Language processing, speech recognition, face recognition that is written in Python, PHP and Objective-C.
I know the interest is irresistible to this point. So, let us waste no time and learn about the pool of things Jarvis will provide to the world of IoT.
Basic Things: Connecting home
Jarvis is cordially designed by the codes that can connect devices speaking different languages and having a different protocol. Zuckerberg even had to reverse engineer the APIs in order to get to the point, so that he can issue a command from his computer and get the devices on according to the instructions.
Also, as we know that most of the devices are not connected to the internet, so for them internet connected to power switches can be used to turn them off or on remotely. But often this thing is not enough. The founder of FB says that he has used Creston system with lights, Sonos system with Spotify for music, a Samsung TV and connected those with his work system at Facebook.
At Jarvis, special emphasis is laid on the speech recognition so that even the common man can use this to work his home appliances efficiently. In addition to this, it has also been taken into account that artificial assistants like Jarvis won’t work until more devices in the industry are developed with common APIs. If these standards are maintained, it would be fairly easy for the developers to integrate these machines to a common device and make the IoT system work more efficiently.
The Visual Recognition
Scientists believe that about one-third of the human brain is dedicated to vision, and the new age AI devices should understand what is happening in the videos and images. Moreover, moving object recognition is also one of the things to look after while dealing with IoT.
Generally, most of the people look at comparing two random objects, but at Facebook, the developers hold expertise in making AI devices that are capable of bifurcating between the faces of humans. This leads the device to recognize the moving object or human with ease and can also inform about the ill practices happening around.
Also, the camera angles should be kept in check as the AI cannot identify the people from the back of their head. With Jarvis, the system is fairly simple for visual recognition: it is a two-step process in which it runs the face detection if any person comes into the sight of view and secondly it runs the recognition to identify who the person actually is. Once the identification is done, it lets the owner know if they are expecting that person and if he/she confirms, he tells them that they are here.
Such type of AI with the IoT is useful in multiple things and can respond to all the context-free requests without needing to check in the request. This aspect of the AI is immensely needed to inform the user about the happenings around the house and helps the house to run smartly.
The coding of Jarvis is done comprising of two processes, i.e. its prompt communication to the text messages and then translates the speech into text for it to read. The coding is full of simple keywords like “bedroom”, “lights”, on, etc. so as to determine that even the smallest of the work is done with precision. As we know that CONTEXT is essential for AI, Jarvis makes it happen for the devices to work according to guidelines and at the exact place where they are needed.
Jarvis has its own set of database and is very efficient is responding to the openended requests.
The Messenger Bot
Primarily Jarvis was programmed to be operated with a computer, but in order to be connected with it all the time and operate it from anywhere; Zuckerberg did the connection with the phone as well.
The messenger bot for Jarvis is made up of a simple framework and automatically handles many things viz. working across Android and iOS, delivering push notifications, maintaining identity and permission for different people and much more. The user can text anything to Jarvis which will be instantly linked to the Jarvis server and the user will have the option of either texting or speaking.
As the present trend of messaging is focused on texting more than voice, the future of AI messaging cannot be limited to voice only features.
For the voice recognition, the founder of Facebook had to build a dedicated Jarvis app. Messenger bot was fine for all the things, but it was causing fraction while listening and understanding the speech. Therefore, the dedicated app will help the user to communicate with Jarvis in a more frequent and lucid manner.
We know that the challenges are never ending, but the inception of Jarvis is a firm step in the direction of AI and IoT evolution. Mark Zuckerberg believes that in the coming 5-10 years, the AI systems will become accurate and will redefine IoT with their powerful technology and tools. This is the time to understand them and their working by implementing things like speech recognition, face recognition, natural language, etc.