« Top Android apps and games of the week #52
Intel Skylake processors: taking gaming to the next level »

Intel RealSense – aspiring to make drones and cars more intelligent

|December 26 2015 |Intel

"Here's how Intel's RealSense can improve cars and drones as we know them"

The world of technology is ever changing, ever evolving, and Intel® is an organisation which has always been at the helm of affairs for a long time now as far as technical developments are considered. With its RealSenseTM technology, it’s now moving beyond computers (quite literally) and hopes to make vehicles intelligent – namely drones and connected, self-driving cars.

In this article, we take a look at all the bells and whistles behind the technology, and how it can improve drones and cars as we know them.


Drones: eye in the sky

While drones are just getting popular in India, they are actually quite popular in other parts of the world such as the U.S. and China. The drone market is growing at unprecedented rates, with more and more people opting to buy drones as holiday gifts or merely out of curiosity.

Originally designed for the military sector, drones are making headlines within the consumer sector as well. While the initial intention of sales for consumer drones was to cater to a market full of drone enthusiasts and developers, the scope of usage of drones has surpassed its own boundaries – no more is it just a toy.


Image credits: http://blog.parrot.com/2014/12/15/how-to-pilot-skycontroller/

Drones are finding increasing use in specialised applications, such as precision agriculture, mining, infrastructure inspection and disaster response. More popularly, they are on the path of making delivery systems and logistics more efficient, as the recent Amazon delivery drones have suggested.

Their most obvious use is in the form of aerial photography and videography, and companies such as DJI and Parrot are actually designing drones with high-definition cameras, advanced stabilizing systems and flight mechanisms to enable filmmakers and hobbyists alike to get a better perspective of things – that too from the air.

Existing drone technology

As far as the current drone scenario is considered, consumer drones nowadays come in quadcopter or hexacopter forms – depending upon the number of propellers on board. Further, depending on the size factors, they may be classified as normal sized, mini, micro or even nano drones.

Apart from stabilising systems and headless modes (which basically implies that there’s no predefined forward-facing mode for the drone), budget drones don’t have much advanced tech behind them. However, some specialized drones such as those from DJI have integrated GPS systems, gimbals and application ecosystems around them, so they can be controlled either manually or can be programmed to go from one point to another while at a specified altitude.


Image credits: http://www.pcmag.com/article2/0,2817,2490234,00.asp

However, drones still need to be controlled or at least need manual intervention. Ultimately, the aim is to make drones self-flying to improve the scope of usage. And for this, RealSenseTM adds quite some vital factors to the equation.

RealSenseTM: improving flight dynamics

Through RealSenseTM , Intel® hopes to make drones more autonomous and sentient, in some ways. As demonstrated in January at CES 2015, RealSenseTM technology enables drones to sense objects and obstacles within its path and successfully avoid them.

In fact, the presenters demonstrated a fun game they called Drone Ping Pong, wherein they individually rushed at the drone and made it veer its way out of the oncoming obstacle – right towards another person who again rushed at it.


Image credits: http://www.pcmag.com/article2/0,2817,2474885,00.asp

All this without any human involvement or manual controls. An infinite loop of sorts, Drone Ping Pong might have been an amusing feat on stage, but it suggests the immense potential of the Intel RealSenseTM technology. In simple words, the technology adds computer vision to the drones.

RealSenseTM technology within specialised cameras enables drones to detect and avoid obstacles on their own volition. The real-time processing engine has been applied to activities such as facial recognition, spatial calibration as well as emotion analysis, and with drones the focus is to make them as independent of human control as possible.

This occurs through a complex system of image processing and the multiple cameras on board the RealSenseTM -enabled drone, as shown in the image above. The software detects the presence of obstacles and dynamically adjusts the path of the drone in order to avoid it and at the same time continue moving towards the destination.

Scope and significance

The scope of this application is tremendous. We’ve already seen how concepts of self-flying drones such as the Lily Drone and the Zano Drone have awed people. This development can give further impetus to drone applications – mere delivery of packages can evolve into delivery of life-saving medicines or relief supplies to natural disaster hit areas, all without human intervention.

Drones which can contextually detect and avoid obstacles, and can intelligently adapt to the surrounding environments – that is what can be heralded as the ensuing advancement in drone tech, and Intel® hopes to realise that soon by enabling drones with RealSenseTM.

Beyond drones: cars with RealSenseTM

While drones equipped with RealSenseTM definitely present a radical approach as to how drones function, the scope is, at the end of the day, not necessarily affecting the end consumer. So let’s talk about a technology which is far more common, and the outreach is actually on a global scale – cars.


Intel® has also demonstrated how RealSenseTM may be applied to the automotive world, at the Automobil Elektronik Kongress in June 2015. In the keynote, a strategy for a next generation software-enabled car was discussed, which could enable cars to be more intelligent and contextually adapt and understand the driver’s requirements.

Through RealSenseTM, a powerful ecosystem is created within the car that incorporates an advanced car infotainment system and a gamut of cameras. Features such as gesture and facial recognitions to lock and unlock the car according to user profiles set, and profile recognition to automatically adjust small things such as the seat position / height and the mirror alignment to intelligently adapt to the driver of the car become a part of the vehicle.

But more interestingly, RealSenseTM can enhance the way the vehicle interacts with its external environment. The RealSenseTM camera is a three-dimensional one, capable of recording objects in three dimensions and rendering them in 360 degrees on a display.

The real significance of this is realised in the obstacle detection system within a camera, and the parking assist feature – both of which render the vehicle in 360 degrees as it would appear in a third person view. So the driver can actually see how the hindrances are placed around the vehicle relative to the vehicle’s current position, and accordingly manoeuvre the car.


Image credit: https://blogs.intel.com/iot/2015/11/23/building-the-next-generation-car-with-intel-iot/

This technology will become even more significant in the near future, where self-driving and connected cars will become more popular. Companies such as Google, Tesla, BMW and Audi are already developing advanced cars that are capable of sensing obstacles and roads without a human at the steering wheel, and in this scenario, RealSenseTM presents itself as a more viable and effective approach for self-driving cars.

Finishing thoughts

RealSenseTM hopes at blurring the divide between the technology of today and the innovations of tomorrow. Bringing about technological advancements to concepts which hold potential for the future and paving the path for a more intelligent, sentient and adaptive environment is the agenda for Intel®, and concepts such as RealSenseTM will help achieve this vision very soon.

Comments (0) View All