Hacker News Reader

Comment by neurotech1 | Hacker News Reader

One thing about the RPV/Quadcopter debate that is rarely mentioned is the reason why they don't use ADS-B transponders.

The FAA requires ADS-B trasponders to have high accuracy GPS, and that pushes the cost to over $2,000 per device. It would be logical for the FAA to relax the GPS requirement slightly, so a cheap GPS module is sufficient to alert nearby aircraft of RPV activity over a certain altitude (eg. 200ft AGL) These RPV-grade ADS-B transponders could use a limited signal output, to avoid nuisance pop-ups from longer distances. The transponder Mode-S ID uniquely identifies the RPV.

It would be possible for a transponder to use an alternate channel frequency, similar to how many General Aviation aircraft use 978Mhz ADS-B. Even with an alternate RPV channel, the RPV operators would still be alerted to regular aircraft operations.

View on HN

Interesting idea although there are a lot of technical issues that would make what you describe more challenging. A simpler approach is positive enforcement. Force manufacturers to hard code drones to not fly near airports or other restricted areas and not fly above 200 ft.

Model aircraft have existed for a long time and there's been a good relationship with the FAA and few incidents. The challenge is that technology has advanced to the point that these modern "drones" are almost too easy to fly. Any idiot can watch some YouTube videos and then they think they're a pilot. Given that, to the extent possible the aviation regulations should be enforced at the device level to make these things as idiot proof in terms of busting airspace regs as they are idiot proof to fly.

I thought the DJI Phantom quadcopters do have firmware lockout for certain restricted areas. Most RC flying clubs are committed to being good neighbors and not conflicting with other aircraft.

Part of the issue is that Helicopters do fly quite low, well away from airports, and the DJI Phantom can be flown high enough to be a concern.

Another part of these new regulations is identifying the RPV so that appropriate enforcement or education can occur.

Some people have suggested that DJI Phantom and other RPVs are shipped in "idiot" mode with tight restrictions. Once an operator can show responsible operation, then switch to a less restrictive mode.

> Force manufacturers to hard code drones to not fly near airports or other restricted areas and not fly above 200 ft.

A drone is one of the few things that can be smuggled by themselves. Both US-Mexico and US-Canada borders are wide.

>A drone is one of the few things that can smuggle itself. Both US-Mexico and US-Canada borders are wide.

Fixed that for you.

And if they add new airports or classified areas ?

It also doesn't solve the problem of drones being active near emergency situations.

"Real" airplanes have the same issue. That's why all the charts and databases are updated frequently and pilots are responsible for ensuring the latest data is available and loaded into the onboard computers. On a typical small general aviation airplane that typically involves uodating the electronic flight bag app on your iPad and popping a new memory card into the on board navigation computer.

Other short term updates (like firefighting flight restrictions) are reviewed by the pilots on official FAA endorsed websites or by calling a flight service center (like customer support for pilots) before each flight. One of the most basic elements of the required preflight planning for any flight of a "real" airplane is to do all the above to make sure you have the latest data and know about any temporary restrictions.

There's no reason why drone pilots couldn't or shouldn't be required to perform a simplified version of the above before flight (if you want to do anything other than flying the thing under the treeline in your own backyard).

A "drone region lockout" bypass (using spoofed GPS signals) was showcased by a team at Def Con this year. They showed video of their drone taking off inside restricted airspace in Beijing. In other words, if someone wants to deliberately break the rules, they're going to find a way. I guess it would still prevent accidental airspace violation.

My main concern is that the definition of "restricted airspace" is likely to increase if this becomes a standard. E.g. the National Park Service would probably lobby to have all of their airspace included, even though it's a nuisance to some people in national parks instead of a safety concern.

There is nothing about aviation GPS's that make them high accuracy. In fact they're mostly inferior to consumer GPS equipment.

What drives up the cost is certification of the equipment.

Certification and also a lot more technology behind figuring out when the position data is inaccurate. Aviation GPS units will run forward looking simulations (for example when flying a GPS landing approach) and figure out that satellites are about to be in a suboptimal position and warn the pilot accordingly.

That's still a simple thing to do with software. The cost of the device shouldn't go up because of that.

That makes no sense, unless your software guys work for free. The cost of software development has to be paid for out of the price of the devices. Aviation is a very limited market, so you don't have a lot of sales to amortize those costs.

The sales volume of consumer GPS units, however, is extremely large; if there is any pure software improvement to be made, even if the benefit for those devices is quite low, I'd expect it to be commonplace. Does anyone have a citation for aviation GPS using more sophisticated algorithms than consumer GPS?

Aviation stuff isn't necessarily an improvement for consumer units, though. For example, an aviation unit used for precision approaches or collision avoidance needs to detect when accuracy decreases below minimums, alert the pilot, and disable the functionality. For a consumer unit you just keep pumping out inaccurate, best-effort numbers.

Most GPS devices I've used provide a precision estimate already.

How accurate is that precision estimate, though? Is it built with enough care that you can put lives on the line?

Appeal to fear isn't an argument. Either the precision is right or it isn't. Precision requirements aren't unique to aviation and aviation doesn't even have the strictest requirements. The required precision of surveying equipment is much higher than what aviation needs.

I would have much more faith in an open source project with lots of people independently verifying the precision of calculations for known devices than closed source stuff from aviation companies that hide behind fear rather than real justifications for the price.

It's not an appeal to fear. It's an appeal to testing and verification.

The cost of these devices are much greater than the cost to implement the feature we are talking about though (alerting on bad precision). You can argue that the cost should be high because of intelligence about runway approaches, but just knowing that the precision is bad is not an excuse.

From a practical perspective that may be true.

Consumer GPS (eg. Phone/Tablet) is typically augmented GPS, whereas certified aviation GPS uses WAAS (Wide Area Augmentation System) to improve accuracy.

The key difference is that aviation GPS will warn the pilot if there is degraded performance below requirements, whereas consumer GPS doesn't.

Here is one ICAO paper on ADS-B accuracy: http://www.icao.int/APAC/Meetings/2011_ADS_B_SITF10/IP10_AUS...

Aviation GPS typically have a greater update rate, and support WAAS (wide area augmentation system) and RAIM (receiver autonomous integrity monitoring).

Just how accurate is this 2000$ GPS supposed to be? Because my 40$ Ublox M8N GNSS modules on several copters produce a solid lock with constant HDOP<1.5 under normal conditions, averaging around ~1.5m horizontal position deviation, and altitude is locked in via barometer down to several cm. To get more position accuracy, you really need DGPS, which AFAIK planes don't normally use either... and do those couple meters really matter?

Its about certified accuracy. The UBlox or similar module could drift under certain conditions.

From a practical point of view, knowing where the RPV is within a 100 meters would be more than enough.

The $2000+ GPS is certified to provide position under Instrument Flight Rules for an RNAV (Required NAVigation performance) approach and landing. Not required for a RPV flown under 400ft visually.

>that pushes the cost to over $2,000 per device.

Sounds like a market that is ripe for disruption. $2,000 cannot be remotely close to the cost of production. This price is the result of having a small, wealthy customer base that is required by law to purchase it. If someone introduces one with pricing that reflects both the cost of production and a dramatically expanded customer base, that company will make a ton of money.

Just like medical devices or military hardware don't cost X to manufacture the cost is in the degree of certification required. Your MVP might look like it does all the same stuff but it is about proving things like mean time between failure, climatic conditions it can operate in etc.

In this case, any up-front costs, such as certification and R&D, can be amortized over each unit sold in the dramatically expanded market. If the market suddenly goes from a few thousand units per year to a few hundred thousand, the contribution that each unit sold must make to recoup up-front costs gets very close to zero.

The cost isn't only about "production" per se; the increased cost also comes from meeting regulatory, reliability and testing requirements.

That is a fucking sane and logical idea, no wonder no one has mentioned it.

Reply on news.ycombinator.com