Three ways in which the video revolution impacts IoT – Embedded

Advertisement
Embedded.com
Embedded vision doesn’t just mean adding a camera to IoT. Instead, it demands an all-in-one system that captures image streams and processes the data on-board.
Advertisement
Connected video devices are everywhere. From doorbells to cars, cheap image sensors from old mobile phones are finding new life in connected devices. This marks a significant transformation in the way we create and consume devices.
As a result, video is moving from a primary feature to a secondary attribute in the internet of things (IoT). These cameras aren’t just altering the way we perceive video – they’re revolutionizing consumer demand and embedded device development in three major ways.
Thanks to this evolution, users now expect even the simplest of embedded sensors to record, process, and transmit video. A connected oven, for example, no longer just measures temperature but contains a live video feed to monitor the dish. Moreover, some models even use artificial intelligence to view the image, gauge the cooking process, and suggest remaining times. This is many more times complex than a basic embedded design, posing a steep learning curve for developers transitioning to video-centric systems.
It’s worth noting that offering embedded vision doesn’t just mean adding a camera. Instead, it demands an all-in-one system that captures image streams and processes the data on-board. In more advanced cases, as touched upon, the device can leverage other technologies to make autonomous decisions. It’s a specific system made for a specific task.
As developers quickly find out, creating this kind of device is only viable with specialized components. For example, real-time operating systems may struggle to efficiently process and transmit video. Often, this requires switching to a high-end operating system like Linux and its library of adaptable software, though this can have energy considerations for battery-powered devices.
Additionally, specific combinations of chipsets are usually necessary for efficient video encoding and compressing. And, since video multiplies the data volume by many factors, developers will need other hardware add-ons to encrypt the imagery for transmission. In all, creating video-capable devices is no mean feat and results in software, hardware, and development obstacles.
It’s for this reason that device makers should think long and hard before taking the video plunge. Not only will they need to redesign the hardware and software of their devices, but they will also need to (safely) create companion apps and platforms to connect users with devices. This requires a special focus on data transmission, security, and privacy. As such, flexible teams with versatile skills will be best placed to answer this demand.
However, not every company will want or need to create video-enabled ovens or doorbells. Instead, leaders must evaluate this evolution in the context of their products. This includes prioritizing research and development (R&D), considering the use case, and adapting accordingly. Most importantly, don’t just add a camera for the sake of it. Trends come and go. Only evolve in this direction if it makes business and technological sense to do so.
Also, remember that device vision is not always synonymous with cameras. Lidar sensors, for example, can determine distance by targeting an object with a laser and measuring the return time of the reflected light. Driverless cars use this technology to map a three-dimensional model of the environment. In some cases, offering this sense of “vision” is easier and cheaper to integrate than an embedded camera. Further, in terms of privacy, customers likely prefer a sensor instead of a camera inside their home.
Of course, adding cameras to everyday devices opens up a security can of worms, particularly when these devices often have limited defense mechanisms. For example, during the Mirai botnet attack, a Chinese surveillance manufacturer recalled almost 5 million cameras with guessable default passwords.
Therefore, this isn’t a device redesign to take lightly. Authenticating users, for example, must be top of mind. Tools like public key infrastructure (PKI) make this possible by employing cryptographic keys in lieu of traditional usernames and passwords. This provides two benefits. First, an exceptionally high level of resistance against brute force attacks. Second, the elimination of shared secrets such as passwords. The latter point is especially important in the event of a security breach or internal compromise, which demands a complete reset of all deployed devices.
It’s also crucial to securely transmit the processed data to the end user, whether through direct peer-to-peer connections or cloud-based solutions. The former bypasses any middleman server to guarantee end-to-end communication. The latter is another popular transmission method, though it can impact affordability since cloud storage providers charge by file size (a valid consideration in dealing with video which produces a lot of data).
As devices invade the modern home and office, increasingly with added cameras, developers must carefully consider their next steps. They must prioritize innovation that makes both business and technological sense, all while safeguarding user data with robust security measures. The video revolution is here – it’s now up to device makers to chart their way forward.
Carsten Rhod Gregersen is CEO and founder of Nabto, a P2P IoT connectivity provider that enables remote control of devices with secure end-to-end encryption.
Advertisement
You must Sign in or Register to post a comment.
This site uses Akismet to reduce spam. Learn how your comment data is processed.
Advertisement
Advertisement
Advertisement

source

About The Author

admin

See author's posts