In the land of Oz, the Tin Man wanted a heart, the Cowardly Lion wanted courage, and the Scarecrow wanted a brain. A hundred and plus years later, tech companies such as Apple, Google, Microsoft, and Amazon have sought to endow their personal assistants with a soul – and, in the case of Siri, some sass. NAR, meanwhile, is working on bestowing intelligence upon drones.
NAR, short for Next Automated Robots, is an up and coming startup from Beirut that is developing smart drones. NAR does not manufacture drones. Instead, the company has chosen to supplement existing models with a processor, software, and sensors if need be, for the purpose of hastening certain processes.
“The drone is a platform”
The idea belongs to Charlie El Khoury and Nicolas Zaatar, both 23 years of age, who had initially devised it during their last year of college for their final year project. Typically, drones mechanistically collect data and relate it back to an earthbound system. NAR however sees the marriage of drones and onboard software as the true essence of the technology; “The drone is a platform,” proclaimed El Khoury. The primary benefit of their approach, he says, is the ability to produce real time notification in critical, time sensitive situations.
My initial impression was that this new paradigm is hardly any different from what’s currently achievable. Shouldn’t data collection, transmission, and processing be exactly the same as collection, processing, followed by the transmission. All the elements are still there; we’ve merely flipped two around. Not so. Because in the first scenario, transmission is considerably more time consuming. Since processing happens on the ground, the drone has to transmit sizable chunks of data, particularly when photographic or video content are needed. Typically, to decrease the load, the quality of the content is reduced. Transmission becomes faster, but the tradeoff yields lower quality data, which, in turn, might lead to lower quality or even faulty analysis.
However, if the processing were to take place on board the drone, then the data transmitted would be substantially smaller. Take the case of a drone monitoring an oil pipeline, the message could be: “leak detected; size: medium; location coordinates: X,Y”. Mere bytes.
With that in mind, Khoury and Zaatar mulled over their target market. They narrowed it down to industries which involve sizable expanses that are hard to access or too large to monitor effectively; industries where risks are high and severe and where timely intervention would be critical. Ultimately, the pair opted to focus on the oil and gas industry and develop software capable of detecting spills and leaks. For this purpose, the drone has been equipped with an image sensor and an thermal sensor.
NAR is keen on collaborating with drone manufacturers. The company is in contact with a number of manufacturers from France, Australia, the US and Canada, whose response has been positive so far according to El Khoury. Discussions have typically revolved around the technical aspects of the machines on offer: the types of sensors on board, the payload capacity, and the physical design of the drone.
Dabbling with hardware
While NAR’s focus and forte lie in software, it does dabble with hardware: Some of the models on offer already come equipped with sensors, but they’ve had to add some themselves in some instances. At times, manufacturers have been open to perform minor tweaks on their designs to accommodate NAR’s requirements too.
El Khoury did concede however that they entertain the idea of building drones themselves. As a small startup, they’ve strategically chosen to concentrate on software for the time being, but he assured me that the passion and ambition to create their own machines are there. If all goes to plan, El Khoury estimates that their hardware venture could begin in about three to five years’ time.
There are several benefits to building your own hardware. First and foremost, it would mean that you are no longer dependent on other entities, which entails fewer chances for unexpected changes and upsets. If a supplier decides to withdraw for example, then NAR would have to rework their design to fit a new model. Making your own drone also means that you are no longer forced to compromise with your suppliers, who oftentimes have different priorities when it comes to the design and performance of the drone.
NAR, which officially launched in September of 2015, has so far undergone three acceleration programs. After winning the Microsoft Imagine Cup in May in that same year, it enrolled in Beirut’s SPEED accelerator, which culminated in a trip to Silicon Valley where it came in second in its demo day. That was followed by a two months at Blackbox Connect in California, before finally returning back home, incorporating the company and joining the Berytech incubator where NAR currently resides.
NAR expectes its first commercial product to be available in 3 months. Beyond that, NAR is aiming far beyond oil and gas, as its model can be replicated in a number of suitable industries, including wildfire monitoring, agriculture, rescue, surveillance, shipping and traffic monitoring.
The post A Beirut-based Startup is Developing Smarter Drones appeared first on MIT Technology Review.
مدونة المستكشف - Bh4b.com