- This article is filed under:
- » View All
Air, Sea, Land: Drones Are Moving Target for Machine Vision Technology
by Winn Hardin, Contributing Editor - AIA Posted 12/14/2015The first comprehensive FAA regulations for commercial deployment of drones, or unmanned aerial vehicles (UAVs), are expected to go into force by 2017 in the U.S. But commercial companies aren’t waiting on regulations to start developing their industry.
Truth is, UAVs and drones have been operating commercially around the world for some time, and the market is gathering steam. Today, dozens of U.S. companies are flying drones for commercial applications under special exemptions from the FAA to do jobs that would be more dangerous without the use of drones, such as inspecting the undersides of bridges and elevated power lines.
It’s for these reasons and more that companies such as Amazon are looking at last-mile drone deliveries and why tech billionaire Mark Cuban is an investor in Percepto, a supplier of computer vision development platforms for unmanned drones and similar applications. Combine these complementary interests and it adds up to a pile of cash for commercial UAV/drone applications. Market analysis firm Business Insider predicts that by 2024, the combined civilian and defense markets for aerial drones will reach $13 billion per year, with commercial applications representing more than $3 billion, or nearly 25% of the total market. That’s compared to less than $500 million for civilian applications in 2013, a fraction of the total $5.6 billion global market for the same period.
Given the success of global powerhouses like Amazon, and the press coverage surrounding founder Jeff Bezos’ interest in using drones to deliver packages, one would expect package delivery to be the killer app that will lift commercial drone applications well above their military predecessors. But that’s not the case, according to Business Insider; rather, the global commercial drone market “will take shape around applications in a handful of industries: agriculture, energy, utilities, mining, construction, real estate, news media, and film production.”
According to the proposed FAA small unmanned aerial systems (UAS) rules, “in the near term, commercial drones in the U.S. will most likely operate below 500 feet, with line-of-sight between the drone and the operator, and weigh less than 50 pounds,” says Keith Wetzel, director, new product development at IMPERX Inc., manufacturer of high-performance cameras for defense and machine vision.
For these types of payloads, cell phone camera technology with small optics, low power consumption, good resolution, and low unit costs will be the technology of choice for most imaging systems on man-portable drones, says John Butler, manager, sales operations for video interface specialist Pleora Technologies Inc. (Kanata, Ontario, Canada). “We’ve seen success in larger-platform drone applications where the imaging data is the key performance driver, such as surveillance or inspection. While it does not have to be a Predator-size UAV to take advantage of Pleora’s interface technology, typically there’s less demand for uncompressed real-time image transfer in smaller form-factor UAVs for consumer or commercial purposes.
“Where we have had success is inside the gimbal, where customers want to use GigE to simplify cables,” Butler continues. “Outside the gimbal, we have frame grabbers that will convert older analog signals to GigE for processing of digital data on-board, archival of raw image data, and transmission to ground stations.”
IMPERX’s Wetzel expects that an FAA ruling on medium-altitude commercial drones will take a long time to finalize, and the opportunities in larger unmanned aircraft may take 10 or more years to materialize. “For smaller UAVs, we plan to introduce a ruggedized camera that is smaller and more cost-effective than our Bobcat line of cameras and expect it to find a home in some of the more demanding small UAS applications,” Wetzel says. “The hottest emerging application for imaging in unmanned systems is for collision avoidance, and that goes across all three application areas: air, sea, and land. When it comes to air and sea, sense-and-avoid systems are more straightforward because the environment is homogeneous and not complicated. On the land, however, the environment becomes a lot more complicated and will present some daunting challenges.”
As driverless cars ride out of the movie screen and onto public roadways, machine vision may have a new market opportunity. “Our products have been designed into automotive manufacturing applications for quite some time, but in recent months we’ve started to see the opportunity for GigE in driver assistance, telematics, and eventually driverless car systems,” says Pleora’s Butler. “Cars have been using Ethernet for a long time, but of the slower 10/100 variety, and automotive electronics are often managed as individual domains with independent controls and processing systems. Driver assistance and self-navigation applications parallel what we’ve seen in windowless military vehicle design, where GigE provides a framework to integrate multi-camera and sensor systems used for navigation and surveillance. Now, instead of one or two cameras in a car, you need the ability to see 360 degrees and know what’s coming at all times from all directions. On top of that, uncompressed imaging and sensor data needs to be more easily shared in real-time across different platforms in the car.”
Whether they come from Google or BMW, General Atomics’ Predator, or DJI’s Phantom 3, unmanned vehicle markets measured in the tens of billions of dollars hold lots of customer segments, each with its own opportunities. Although today’s consumer drones will not likely adopt industrial-grade imaging technology en masse, high-value systems for bridge and high-voltage wire inspection may very well need the ability to quickly adjust dynamic range while accurately measuring the spatial characteristics of hidden structural defects. In other words, the question remains: Why would a $1,000 drone need a $4,000 camera system? Perhaps the bigger consideration is this: Is that $10 billion bridge worth a high-quality image, or, how much would you pay to keep your children safe on the way to school as the wheels on the autonomous bus go round and round?
There are currently no comments for this article.
Leave a Comment:
All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.