This thorough answer by @Hobbes links to the item Auto-destruct system seen as a key to ramping up launch tempos.

I'd like to understand Why would autonomous auto-destruct ramp up launch tempo?

I have tried to read through it, and I think the answer is in here, but I'm not familliar with even the basics of range safety. Could someone help me understand better what the "resources on the range" might be, and which "antenna pointing at the vehicle" is no longer needed? I'm sure there are still some antennas somewhere pointing at it.

“We have this problem right now where we negotiate dates back and forth, but I think it’s actually going to get better,” Koenigsmann said Feb. 8 at the FAA-sponsored Commercial Space Transportation Conference in Washington. “The reason it’s going to get better is the Autonomous Flight Safety System. It does not need a lot of resources from the range, so the time it takes to reconfigure the resources from a ULA vehicle to a Falcon 9 or some other vehicle, that factor will basically go away.”

While the range is still involved in launch operations, “there is not an antenna pointing at the vehicle, so that in itself, I think, will make our lives easier,” Koenigsmann said. “I could easily imagine that we’ll have two launches on the same day because of that.”

According to the article, Hans Koenigsmann is vice president of flight reliability at SpaceX


3 Answers 3


I haven't found a full list of range activities, but there's a key comment in this article:

The Air Force currently requires several days to reconfigure its ranges between Atlas, Delta and Falcon missions. That turnaround time should be reduced with the introduction of auto-destruct mechanisms

So in the old situation, every user had its own self-destruct system that needed some unique equipment or configuration on the ground (because the system expects information in a unique format, for example). Because the self-destruct system is, well, destructive, they'd want to test the system after every reconfiguration to make sure it works and that no false destruct orders will be sent. That alone could take a few days.

During the launch, they'd also need to provide the Range Safety Officer with all the information they need for the decision in real time, again unique information for every launch vehicle. The RSO is the person with the finger on the 'Destroy' button.

In the new system, the flight termination rules are decided beforehand, and an automatic system on the rocket checks if the rules have been violated. That means you no longer need a ground-based system to receive and analyze telemetry, and translate the data for human consumption in real time.

It turns out, that ground-based system was really manpower-intensive:

As the Brig. Gen. explained, “We have now gone completely autonomous with that system. So with CRS-10 and all others with the AFTS, we’re able to reduce our operational footprint by 60% on day of launch."

“So we came down 96 people that don’t have to be sitting on console. And the cost to the customer is cut in half."

Those 96 people were involved in:

Comm, radar, transmitter, receiver, backup power generation, software, tracking cameras, console maintenance, etc They would be located at the MOCC, JDMTA, Antigua, Cape command antenna site, camera sites, etc.

  • $\begingroup$ I don't think this part "They'd also need to provide the Range Safety Officer (the person with the finger on the 'Destroy' button) with all the information they need for the decision, again unique information for every launch vehicle. " is going to change. Hence the quote "We still develop the mission rules to provide public safety, but the system works with mission rule data files loaded into the on-board AFSS units. This essentially shifts the workload to the front-end of the launch process.” The AF is still responsible. Your part about the system reconfig & testing is spot on. $\endgroup$ Commented Mar 15, 2017 at 13:30
  • $\begingroup$ OK, (until now) each system is different, complex, distributed, includes both human and electronic elements, and critical that it function correctly, with no false positives or negatives. I can certainly imagine a significant amount of testing and qualification would be needed each time. Standardization almost always brings improvement in throughput and reliability (in general at least). Thanks. $\endgroup$
    – uhoh
    Commented Mar 15, 2017 at 21:34
  • 4
    $\begingroup$ It sounds like most of the advantages have nothing to do with the system being automatic, but rather with the system being standardized, which apparently it wasn't before. Or from a slightly different perspective: removing the need for it to be standardized since it isn't actually interfacing with the outside world anymore (except for the format of those "data files" mentioned, which can and probably will be converted to a non-standard, flight-computer-specific format before uploading them to the rocket). $\endgroup$ Commented Mar 16, 2017 at 11:08
  • 1
    $\begingroup$ The new system is only partly standardized. Part of the hardware and part of the software is standardized. The vehicle vendor has to provide vehicle-specific hardware and software, plus vehicle-specific and flight-specific flight termination rules. The standardized and vehicle-specific hardware and software can be verified and validated well before launch, moving that part of the process out of the launch processing. The flight-specific termination rules are written in terms of the automated equipment. This can also be pushed up earlier, presumably prior to flight readiness review. $\endgroup$ Commented Mar 18, 2017 at 18:40
  • $\begingroup$ @OrganicMarble: i've rephrased that part of my answer. $\endgroup$
    – Hobbes
    Commented Mar 21, 2017 at 14:54

Basically, the short version of range safety is that there's a network of systems that all come together. Each system has a number of people that are required to physically be at the launch making decisions.

So for example there's telemetry, radar, etc. AFSS cuts the number of people who have to physically be at the launch significantly down (so this allows for people to not have to work multiple launches per week/day, etc.) The last point is important because there's rules in effect to how often a person can be at work, how long before/after a launch they can work. By eliminating this, you have effectively have removed that limit.

But, and more importantly, the range instrumentation is eliminated, mostly. You don't need all the radars, telemetry sites, etc. AFSS uses GPS and an on-board telemetry system. So you don't have to take maintanence of the range, etc. into account. Additionally, you've removed some points of failure (so for example, if you don't have a radar, you don't have to worry about that radar failing).

So in essence, you're reducing number of required personnel, taking away range assets, thus allowing for a faster "reset" time on the range. And that's what allows for an increased launch tempo.

Lastly, because you don't have to have personnel in the range safety rooms, it allows you to be able to have multiple launches on the pad simultaneously, allowing for several launches to go up, even as quickly as on the same day!


In my limited work with the big ranges (e.g. VAFB or CCAFS) a machine is definitely going to be faster. That's not throwing shade at the dedicated people who work there, only that each rocket, nay launch, is treated as a special case. With an autonomous flight safety system (AFSS), all the coordination is done up-front, once, and then not revisited with every launch. It’s really no different than the value of automating any process. There's an up-front investment paid back as an annuity thereafter.

  • $\begingroup$ While what you say may be true, can you focus a little bit more on the question - what is it about launch tempo that would get faster? What does your term "coordination" actually mean in this context? I'm asking this just to try to improve the "stackexchangedness" of your answer. Thanks! $\endgroup$
    – uhoh
    Commented Mar 22, 2017 at 4:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.