Robot “dogs” to keep USAF runways free of dangerous debris


A robot described as “Roomba for the runways” under development for the US Air Force could help prevent foreign objects from damaging planes.

Foreign object debris or FOD – anything not supposed to be on the tarmac – causes the aviation industry an estimated $ 4 billion in losses each year. The damage from FODs can be catastrophic, causing Air France flight 4590 Concorde to crash immediately after take-off, killing all 109 people on board in 2000.

FAA regulations require airports to perform regular FOD inspections. In the military, runways and flight decks are frequently visited by FODs with available personnel called in to walk the flight line and check for potentially dangerous objects.

The danger is twofold. Any debris can puncture a tire or be ingested into an engine, and can be important evidence of other problems as well. If an airplane is missing a nut, bolt, fastener, or other component, it can lead to more serious problems.

Many airports have sophisticated radar systems to detect FODs, but these are expensive and do not necessarily provide complete coverage. Walkdowns and drive downs are therefore still a daily reality.

New technology being developed for the Air Force aims to change that.

The four-wheeled robot known as the FOD Dog is being developed under a US Air Force contract signed this month. The Siemens Government Technologies division is the prime contractor, with TurbineOne subcontractors working in the collection system. The contract is for one year, with a fully functional prototype system to be demonstrated by next January.

The robot itself is a basic rugged chassis, with an electric motor, and it scans the tracks for foreign objects using standard laser sensors. These detect the profile of the soil strips, detecting anything protruding from the surface. The advantage of the scanner over a camera is that it is simpler and works equally well in poor visibility conditions such as driving rain.

However, the real magic is the technology that makes sense of the scanner data to detect and identify FODs. CEO of TurbineOne Ian kalin – who participated in the FOD field trips while serving in the US Navy – says there are three enabling technologies behind it.

“First of all, there are graphics processor units (GPUs) from companies like NVIDIA

. These rugged little cards are the new industry standard for video processing, ”Kalin told Forbes. These are the units that advance small drones and other mobile robots.

“Second, machine learning models have gone through waves of ‘miniaturization’, which essentially means that code that previously required large computational resources and many gigabytes can now run on just a few megabytes of space,” explains Kalin.

This means the latest intelligent machine learning software can run on small, cutting edge computing devices rather than server racks like earlier machine learning projects.

FOD dogs start with a catalog of different items, but are able to learn new ones. They can even share information among pack members, so they will gradually become better at identifying specific types of FOD at a particular site. This can be important when a site receives a specific type of foreign object, such as leaves.

“Finally, TurbineOne invented a front-line perception system, which serves as an operating system to make all of these detections possible without an Internet connection,” says Kalin.

This allows the dog pack to work in places where there may not be an Internet connection, such as military sites or aircraft carriers.

FOD visits are generally carried out at night after the end of flight operations or early in the morning. Each of the robots automatically plugs into its charging station (“dog kennel”), much like a household cleaning Roomba, and the dogs will sweep, working together to provide full coverage. A human operator can manage an entire pack.

Any localized FOD is reported automatically. Robots can either pass the exact coordinates to their operator – the improved built-in GPS means they locate it within inches – or a robot can physically lead the operator to the FOD like a hunting dog.

Kalin says debris removal is not a requirement for the initial contract, but they have it on their roadmap. Their likely solution will be to equip certain vehicles with a standard robotic arm and manipulation software. This would turn the robots into a complete track clearance solution.

TurbineOne sees many other uses for this technology beyond FOD detection. Drones are already used for industrial inspection, but have limited flight time and cannot fly in bad weather. Ground-based robots have longer endurance, can ignore weather conditions, and carry a heavier payload of various sensors. TurbineOne recently announced $ 3M in seed funding from XYZ Venture Capital to expand its engineering team and accelerate product delivery.

Robots equipped with perception systems could perform many inspection tasks – perhaps starting with inspecting aircraft, similar to a drone-based solution being deployed in Korea, but extending to factories, to construction sites and other applications. (Perhaps the robotic inspection robot could have prevented the recent F-35 crash apparently caused by a rain cover left on an engine).

The combination of a robust and inexpensive robotic platform and intelligent sensing has great potential and the FOD Dog can be the forerunner of many useful robot breeds working in a variety of fields.


Comments are closed.