Wednesday, May 31, 2023

Robots Invade 6 Meters

The return of 6 meters sporadic E marks my return to digital modes. With rare exceptions, my operating on digital modes is greater than 95% on 6 meters and the rest on 160 meters. I do the former for its superiority at exploiting rare and fleeting DX openings, and the latter when there is little CW activity to be found.

FT8 in particular is wildly popular. Since I have little interest in digital modes, I use the technology where it fits my interests and otherwise stick to CW and SSB. It's a matter of personal preference. I don't look down on those who predominantly operate digital modes. That's an attitude many hams of my generation regrettably profess.

Many of those attracted to digital do it to compensate for poor antennas. In the past that would drive them to CW, which is not an easy option for recent licensees that may not know CW. Others do it because the technology is fascinating. I would not be surprised to learn that some use digital to avoid unwanted conversations. All are valid reasons, in my opinion, even if mine are different.

It is no surprise that fully automatic -- or robot -- operation is seen on FT8. The software-driven technology is eminently suited to it and there are ample hams with the expertise to do it. After several years of there are numerous forks of the open source WSJT-X software that include automatic features.  All you need to do is download one of them and you, too, can be an FT8 robot.

Should you? Is it wrong? Automatic operation is certainly discouraged by the original software authors and operating award sponsors. Nevertheless it is popular and becoming more so. It is fairly easy to identify robots but difficult to prove with an acceptably small probability of false positives.

It is not my intent to write another critique of digital robots. There are enough of those, and I've spoken of it before. The phenomenon is of interest to me since as more and more hams migrate to 6 meters -- a passion of mine -- and the overwhelming popularity of digital modes, it impacts me every day during sporadic E season. It is worth pondering their prevalence rather than reflexively attacking robot operators.

As a 6 meter DXer, I have several issues with the presence of robots:

  • They occupy spectrum. Robots typically are configured to operate non-stop. The 3 kHz FT8 and FT4 windows can support a large number of signals, but there is a limit and we've reached it during many sporadic E openings. Not everyone will move to 50.323 MHz during hot DX openings.
  • They will call you, regardless of whether you call "CQ DX" or other target area. They will follow you around the band, often calling on your transmit frequency, thus QRMing many callers. They may continue this behaviour indefinitely, even after you stop transmitting.
  • Robots typically call anyone next in line according to algorithm and configuration, without respect to distance, grid or country. They fill their logs with what I consider fluff.

From my observations, it seems that those who run robots are not very good at configuring the software! A few simply may not care that they are an annoyance and causing QRM. However I don't believe that they're evil. I know a few of those who are almost certainly running robots on 6 meters and I would not consider them to be "bad" people. There are moral and ethical questions that have no clear answer. That I and many others dislike robots does not make them objectively "wrong".

Do you want to work them?

Apart from the issues addressed so far, is there anything objectionable about working a robot? For award and contest purposes, it is at least discouraged and will usually result in disqualification of the robot operator, if they're found out. Some are obvious but all are difficult to investigate. That can be a trial for award sponsors that prohibit robot QSOs. That is not my problem since I do nothing wrong by working robots.

I don't call stations that I know or strongly suspect is a robot. When they call me, I may work them despite my dislike. I sometimes do it just to be rid of them. Depending on my mood I may not log the QSO. In most cases I ignore them. There exist ports of the WSJT-X software that have black lists and related features for those that want to automatically ignore suspected robots.

If I'm in a mischievous mood I may toy with robot callers. I will change frequency, switch between odd and even periods, make one response (or one that is deliberately out of sequence) to see what they do. But only if I have a few minutes to waste. Probing their behaviour can help to better identify other robots. 

Let's try an extreme hypothetical situation. Imagine that the robot is in North Korea (P5). Is that enough to convince you to work it, whether it calls you or you answer its CQ? How about a robot dropped on the Moon by a future space mission or riding on the ISS or other satellite?

There is no right answer to these extreme cases. It is worth taking a few moments to consider what you would do. All could occur eventually.

Why do they do it?

No, I don't know why since I have spent little time trying to find out. In one case that I remember well, the ham shrugged his shoulders and said, "why not?" Maybe they don't know either. Does knowing why even matter? There could be as many reasons as there are robot operators.

As I said above, I don't think robot operators have evil intent. The technology makes it easy and it is admittedly interesting. I remember discussing the possibility of automatic CW operation with a friend of mine about 30 years ago. We decided it was possible but difficult, and eventually it would probably be done. Neither of us was interested in following through and it seems few others have bothered. 

With no QRM, QRN and QSB, and well formed characters, robotic CW operation can be done, though not without glitches. It isn't possible to anticipate every response to a transmission. Of course the CW robot could emulate a human operator who wants out of a conversation by forcing an end. A CW robot would have to be rude by design. Digital QSOs are far more structured and thus amenable to competent robotic operation.

One topic that I don't recall discussing with my friend was ethics; that is, the appropriateness of a CW robot. This is perhaps unsurprising. We were both software professionals with a love of CW, and beyond the technical challenge we ultimately thought the result would be quite boring. Why hand over operating to a robot? We want to be the operators! That's why we became hams.

With that limited insight, I doubt that most FT8 robot operators are overly concerned about its ethics. They stay quiet about their activities since they know others would question their ethics. 

Robot operation may be interesting, but a brief passion that is soon discarded. Operating is more fun to do than watch. But if for whatever reason you want to fill the log with contacts and your life keeps you out of the shack, or you want the contacts without the "tedium" of operating, a robot might be tempting.

Where next?

One curiosity about the robots I've identified is that they have clean signals. I can't say the same for not a few of the 6 meter stalwarts! That's likely because they don't use amplifiers and run their rigs conservatively. The duty cycle of FT8 is hard on transmitters and amplifiers that are run non-stop close to the equipment power capability. Robot operation may be annoying and unethical but robot operators aren't stupid.

That is good since robots aren't going anywhere. I expect their number to grow, though slowly and it will likely plateau. While many may at first be enticed by the technology and QSO potential, it is ultimately boring. Individual robot operators will gradually scale back or give it up entirely. Peer pressure from local hams can also be effective. Shaking your fist at the computer display is not effective.

I'll close by noting that I have not mentioned any robot FT8 applications. They're out there and easy to find. I have no reason to make the search easy.

No comments:

Post a Comment

All comments are moderated, and should appear within one day of submission.