df8m1 24 #201 June 21, 2016 I am going to a CRW boogie in Hastings MI this weekend to collect some data and talk to some jumpers so I can see if there is anything that we can do to address their special needs. Sounds like taking a downplane low is the most challenging aspect, but until I get some feedback and data I can't say for sure. Quote Share this post Link to post Share on other sites
jacketsdb23 49 #202 June 28, 2016 I'm very interested in your work. For what its worth, I just installed an Argus on my swoop container with a Petra 68 in it. The Argus is set to swoop mode, which by mfg. definition will disarm the AAD after it interprets an open canopy. I hope this feature is more readily available on future AAD's. Good luck with your testing!Losers make excuses, Winners make it happen God is Good Beer is Great Swoopers are crazy. Quote Share this post Link to post Share on other sites
df8m1 24 #203 June 29, 2016 Thank you for your support!! I think an AAD should be functional in every phase of a jump, not just in one phase because it can not handle another. I also think that we as a manufacturer need to be confident enough in the our AADs ability to correctly identify the flight mode that the jumper is in, that we do not need an audible indicator to inform the jumper that the AAD failed to correctly identify a change in flight mode. Of course it is easy to talk about reaching a high level of performance confidence.... Just food for thought... We are working on the Beta designs now for both the Static Line and Sport AADs. Once those are build and tested in house, we will be sending them to a few jumpers in all the disciplines to collect data that we will use to test our algorithms in simulation. Once the algorithms check out, we will upload them on the Beta units and have them jumped in a passive configuration. Then, after we have enough passive jumps logged, we will rig them live. This will be an exciting process!! I have set some bold objectives to meet and at very high standards. Quote Share this post Link to post Share on other sites
df8m1 24 #204 December 15, 2016 I thought I would update this thread as it has been a while… My focus has been on the platform design and build and I have a bad habit of taking a step back and thinking, “this is nice, but it could be a simpler” , which is what happened and facilitated 2 design revisions.. well one total revision, and a partial regression lol.. My goal was to have airworthy units by Jan 1, but the reworking set that schedule back, I am hoping to have units by PIA. I am not thrilled about the delay, but I am glad I explored the different possibilities, and took the best of them all where I could. I really do not like designing to a schedule as time dictates when the design is done, as apposed to the design being done when it is right. As if the new platform work and PIA preparations were not enough, we are going to Yuma to test our Military Static Line AAD right after the first of the year. This is going to be an exciting new year.. I have applied for 2 seminars at PIA, one on the Military Static Line AAD and the other one on advanced AADs which will be about the Sport and Military free fall AAD that we will be introducing. These AADs are the Next Generation of AADs which will be situationally aware, and able to identify and complement the flight mode of the jumper regardless of whether they are at terminal, flying a wingsuit, swooping, or any combination there of, and without compromising functionality on one end of the jump for another. The AAD will also be able to determine if the jumper is descending at a high rate while still in a plane or if the jumper has exited the plane. This capability is what will prevent the AAD from firing in a plane, regardless of the descent rate or altitude. I am very curious as to how our operational philosophy is received. I have been test jumping our guni-pig when I can and I am very pleased with its capabilities so far. I welcome all comments, friend or foe Quote Share this post Link to post Share on other sites
mxk 1 #205 December 15, 2016 df8m1I have applied for 2 seminars at PIA, one on the Military Static Line AAD and the other one on advanced AADs which will be about the Sport and Military free fall AAD that we will be introducing. I really hope that these get recorded and uploaded to YouTube/Vimeo. df8m1I am very curious as to how our operational philosophy is received. For me, this will come down to whether I understand how the device works and how it can fail. Even though I haven't seen the source code for my Vigil, I have a sufficiently good idea of how it does what it claims to do. Your claims are far more ambitious, which means more complex implementation, and more room for bugs and unforeseen failure scenarios. Once your AAD has been out for 10+ years without any major problems, then you could probably get away with saying "trust me." For a brand new device, however, if I can't understand how it solves the very difficult problem of translating sensor data into an accurate representation of the current situation, it will be hard for me to trust it. Quote Share this post Link to post Share on other sites
uer16 1 #206 December 15, 2016 Sounds very ambitious... How would you distinguish a sitflyer staying in a sit for the whole skydive 'til 1K from a plane diving down from altitude at the same speed to 1K? To me the acceleration and pressure profiles would be programmatically indistinguishable. What hardware design standards do you use? What software design and implementation standards to you use? Static analysis? MISRA? More importantly, How do you model the flight modes? How do you design algorithms that take sensor inputs and spit out what mode you're in? How many states can the device be in, how do you test all possible state paths with 100% coverage? Sounds like an awesome project, and I'm sure it's possible to make it work in 99% of situations, but to me that kind of complexity makes the device generally unverifiable. In life saving equipment you can't beat simplicity. IMO, if(altitude < 900 && speed > 78) { fire(); } Is the way to go. Quote Share this post Link to post Share on other sites
Bob_Church 7 #207 December 16, 2016 How much of the cost of making an AAD goes into making it reusable? I think I'd skip that part. If it fires it saved your life and you have some hard decisions to make, but if you do stick with the sport afterwards it seems like paying full price for another would still be a bargain. Quote Share this post Link to post Share on other sites
df8m1 24 #208 December 16, 2016 I will address the comments starting with mxk I will be videoing both of them and will be putting them on my website even if PIA does not. You are not alone in your thinking. I have had conversations with engineers who, short of me providing them with the source code and algorithm legacy, will forever be leery, if not straight up challenging as to the capability of a black box doing what is clamed and performance testing has shown to work. And that is OK. Unfortunately to satisfy this perspective, the level of disclosure would be beyond what I can provide. We are protecting our algorithms as Trade Secrets and as such, disclosure at any level that could undermine the IP security is prohibited. This is a frustrating part of all this because I really would like to explain in great detail how all the algorithms work, I am very proud of them! It truly is amazing to work with the simulation version and run jump data through it. Watching it come to conclusions is really cool. Please know that although I am a creator, I am also responsible for the business aspects of this technology, which is being considered by the US Government, and potentially other companies who what to have this kind of capability. If I can not careful, I could very easily let a key philosophical aspect get out that could allow a competitor to be one step closer to doing the same. Case in point, Jan 3rd Army Research will be testing our Military Static Line AAD at YPG, along with 3 others from competing companies. I will really try my best to answer questions as thoroughly as possible, but there will be some simple questions that I will not be allowed to answer to anyone’s satisfaction. Please do not hesitate to ask anything. uer16Sounds very ambitious... How would you distinguish a sitflyer staying in a sit for the whole skydive 'til 1K from a plane diving down from altitude at the same speed to 1K? To me the acceleration and pressure profiles would be programmatically indistinguishable. What hardware design standards do you use? What software design and implementation standards to you use? Static analysis? MISRA? More importantly, How do you model the flight modes? How do you design algorithms that take sensor inputs and spit out what mode you're in? How many states can the device be in, how do you test all possible state paths with 100% coverage? Sounds like an awesome project, and I'm sure it's possible to make it work in 99% of situations, but to me that kind of complexity makes the device generally unverifiable. In life saving equipment you can't beat simplicity. IMO, if(altitude < 900 && speed > 78) { fire(); } Is the way to go. uer16, The normally question I get is how will it tell the difference between a wingsuit in flight and a jumper under canopy lol… In an attempt to “carefully” answer your question about sit flying in an airplane… There is a major difference between descending in an aircraft and sit-flying, and that difference is an exit. The first task any electronic AAD has to do is identify if it is on the ground or climbing to altitude, then it looks for the arming altitude, the checks the descent rate against altitude. Our Advanced AAD identifies that is is climbing to altitude as well, and it does look for an arming altitude, but it does not arm until the jumper exits the plane. Without an exit, the AAD is passive. Unlike other Static Line AADs that use a RF signal as a Virtual Static line, or use a physical static line to identify when a jumper has exited, our Static Line AAD is completely stand alone and does not use any other external device in order for it to determine when a jumper has exited. Our Static Line AAD min exit altitude is 525Ft AGL, and in a total malfunction scenario, will fire the reserve 4 seconds after exit. In this application timely identification of the actual point of exit is critical. The free fall AADs exit identification timing will be more relaxed lol… Regarding design standards: Our process is consistent with DO-254 for hardware and DO-178B for Software. Software is not one of my areas of strength, I have learned a lot about it through this process and I can decipher code to an extent if I have to, but I leave that aspect to my software engineer, to whom I will inquire about MISRA. The last thing I want to do is try to pretend I am more knowledgeable in an area than I am. In regards to Flight Mode Modeling and Algorithm design process… We have been working on AAD algorithms for 10 years on and off , maturing, morphing, and creating new ones on a per project basis so we have quite pool to draw from. My basic process is to take data recorded from a known flight, identify what I want the device to do and when. Identify the algorithm(s) that best fit the sequence(s), and rerun the previously recorded data through them and check the result. I have a pool of “Gold Standard” data that we use for QA every time we make a change, be it a logic change or a conditional setting. I really do enjoy that aspect of this process. It can be very frustrating and satisfying, and sometimes both at the same time lol.. Also keep in mind that the US Military is currently vetting our Static Line AAD and we will be putting our Military Free Fall AAD up to their tests as well when it is ready. They have put our Gen-I SLAAD through component reviews, Environmental testing, and Drop testing. The AAD fires into a data recorder so they have an independent record of exactly when it fired. They also have a complex visual positioning system that will provide additional reference to the location of the drop at any given point in time. If our AAD does not identify the exit and fire or not fire in time, they are going to know it. Bob Church: The funny thing is Reusable and Reliable are very closely related. I do not know if it is still available, but there was a hot wire AAD for a while. I believe it was a one-time use kind of thing. The problem is the nature of the application and the volume of production. Another aspect to consider is the legal defense aspect, which is also part of this application. Quote Share this post Link to post Share on other sites
uer16 1 #209 December 16, 2016 Great answers, you seem to have your shit together, wish you success in your endeavors Quote Share this post Link to post Share on other sites
councilman24 37 #210 December 16, 2016 Is the nose over and dive of a turbin aircraft at freefall speeds enough different.from a skydiver exit and acceleration to terminal that the two can't be confused? Mike Mullins' decent has to look .a lot like exit.I'm old for my age. Terry Urban D-8631 FAA DPRE Quote Share this post Link to post Share on other sites
mxk 1 #211 December 16, 2016 df8m1Unfortunately to satisfy this perspective, the level of disclosure would be beyond what I can provide. We are protecting our algorithms as Trade Secrets and as such, disclosure at any level that could undermine the IP security is prohibited. This is a frustrating part of all this because I really would like to explain in great detail how all the algorithms work, I am very proud of them! It truly is amazing to work with the simulation version and run jump data through it. Watching it come to conclusions is really cool. I completely understand your position, and I'm actually glad that you've not gone down the software patent road on this. However, it does leave you the problem of proving that your AAD is better in some ways than the current generation, and that it does not introduce many (any, ideally) new failure modes. SL tests aren't going to be sufficient for this. It's a different and more controlled environment. I'm not familiar with trade secret law, but have you considered doing an independent code audit protected by NDAs? df8m1I will really try my best to answer questions as thoroughly as possible, but there will be some simple questions that I will not be allowed to answer to anyone’s satisfaction. On the software side, the question I'm most interested in is whether the system is based on a machine-learning classifier algorithm or if the core situation identification logic is coded by hand. df8m1There is a major difference between descending in an aircraft and sit-flying, and that difference is an exit. The first task any electronic AAD has to do is identify if it is on the ground or climbing to altitude, then it looks for the arming altitude, the checks the descent rate against altitude. Our Advanced AAD identifies that is is climbing to altitude as well, and it does look for an arming altitude, but it does not arm until the jumper exits the plane. Without an exit, the AAD is passive. Would it correctly identify the exit for all involved in the event of a mid-air collision? What about an accidental main deployment while setting up in the door? Based on what you said about having a pool of gold standard data, I doubt that you have good representations of these two and many other rare situations. How do you know that your AAD will do the "Right Thing" in scenarios that haven't been recorded and simulated? What happens if the system gets confused by unfamiliar, corrupt, or missing sensor data? Quote Share this post Link to post Share on other sites
df8m1 24 #212 December 16, 2016 councilman24Is the nose over and dive of a turbin aircraft at freefall speeds enough different.from a skydiver exit and acceleration to terminal that the two can't be confused? Mike Mullins' decent has to look .a lot like exit. I only have a moment so I have time to address this one lol.. The others I will address later tonight. We have an appointment with a Pitts when we have the new platform up and running. We are going to put in the nose and let the pilot go to town. We will have a cutter simulator that lights up when the AAD fires and a Go-Pro focused on it in the nose as well. The pilot will be exceeding the firing descent rate well bellow firing altitude. Exit identification is a tricky thing, I totally understand questioning that ability. More to come tonight.. Quote Share this post Link to post Share on other sites
Bob_Church 7 #213 December 17, 2016 "The funny thing is Reusable and Reliable are very closely related. I do not know if it is still available, but there was a hot wire AAD for a while. I believe it was a one-time use kind of thing. The problem is the nature of the application and the volume of production. Another aspect to consider is the legal defense aspect, which is also part of this application. " Thanks for the info. Quote Share this post Link to post Share on other sites
df8m1 24 #214 December 17, 2016 MXK First let me say that I am not and will never say that this AAD will be "better" than any other AAD, they all have to meet the same high standards. What I will say is that this AAD will be compatible with all flight modes, from wingsuit to swooping, without sacrificing performance on one end for another. I will meet you half way and say that it will be better suited for the extreme disciplines. The SL testing is strictly for our Militay Static Line AAD. One might think that C-130 or UH-1 static line exits would be consistent, but in reality, the only consistent thing about a military exit is at some during the jump, the jumper has to fall out the door lol…. I mean no disrespect to Paratroopers, have you seen how much stuff they are carrying?? It is impressive.. The SL tests do validate our exit detection and malfunction detection ability, which was the reason for the mention. Software: An AAD is really not a good application for machine learning IMO due every jump is different. There are some sliding scales that do vary some values depending on other values, but all the decision-making is programmed in. The algorithms define condition constraints and look at 10 channels of data in real time for those conditions. That is the simple explanation… Your scenario of a mid air collision (I’m assuming aircraft to aircraft) would not generate the conditions consistent with what our algorithms are looking for that would qualify as an exit, but getting yanked from the door by a deploying canopy would. Have we recorded data of those two events? NNNOOO lol, but I can and have generated artificial data to find the range limits of algorithms. I can create just about any situation with some effort.. This AAD is actually made of two main components, one being the Black Box (flight data recorder) and the second part being a traditional AAD that uses altitude and descent rate. The Black Box commands, confirms, or denies a fire condition. Even though the AAD does not need to be told what type of jump is taking place via selecting a MODE using the interface, such as Wingsuit Mode, it will have modes that can be selected. If a wingsuiter puts the AAD into Wingsuit Mode, it will work just like if it were in Normal Mode, but, since the AAD knows that the jumper is flying a wingsuit, if the Black Box should shut down mid flight in response to a problem, then the AAD will become the master and finish out the flight based on a more traditional speed vrs altitude decision tree, but tailored for a wingsuit. There are fault detection processes constantly running that look for communication errors or bad values, and fault codes are set and stored. Should a fault be detected mid flight, depending on the nature of the fault, the program will either disregard that channel during the remainder of the flight, or revert command to the AAD portion which has it’s own pressure transducer. Should a fault be set mid flight, once on the ground, the AAD will shut down. When the user turns it back on, they will be notified of the fault and the AAD will shut down. The Failure Mode side of this process is intense for sure. And will no doubt be an ongoing process. Quote Share this post Link to post Share on other sites
Divalent 106 #215 December 26, 2016 df8m1... The AAD will also be able to determine if the jumper is descending at a high rate while still in a plane or if the jumper has exited the plane. This capability is what will prevent the AAD from firing in a plane, regardless of the descent rate or altitude. ... Until all AADs have this capability, I don't see the point, as pilots will have to assume that a descent rate that exceeds any AAD in service would present a risk of an in-plane fire. How critical in other areas of operation is this capability? And here's a scenario that I wonder if your exit-algorithm would get correct: for jollies, the pilot does a zero G maneuver, so for a few seconds both plane and jumpers are falling, but in the plane. Exit detected or not? And a variant (that probably is so rare its never happened nor likely to ever happen), during this zero g maneuver, a jumper sitting next to the open door floats out, but early in the maneuver, when the plane wasn't actually falling (it's upward motion was decelerating at 1g). The wind blows him towards the tail, and his head strikes the tail, knocking him unconscious. Exit detected? (So that the AAD saves his life?) Quote Share this post Link to post Share on other sites
df8m1 24 #216 December 26, 2016 df8m1... The AAD will also be able to determine if the jumper is descending at a high rate while still in a plane or if the jumper has exited the plane. This capability is what will prevent the AAD from firing in a plane, regardless of the descent rate or altitude. ... Divalent Until all AADs have this capability, I don't see the point, as pilots will have to assume that a descent rate that exceeds any AAD in service would present a risk of an in-plane fire. How critical in other areas of operation is this capability? You are correct that a pilot under normal circumstances would need to be mindful of the aircraft descent rate because standard AADs could fire bellow their hard deck. That being said, Turbine aircraft can climb to the arming altitude very quickly and an engine out shortly after take off happens now and then. Depending on the aircraft engine configuration, the pilot may very well have their hands full feathering props, and adjusting for the loss of an engine on one side. If this happens just above the arming altitude and the aircraft is difficult to get control of, it is very possible that the plane could exceed the firing descent rate. Some DZs allow the doors of their super otters to remain open on take off, and if some poor jumper in the back has an AAD fire as the plane dives due to a pilots recovery process, then that could make a bad situation worse. Is it a huge problem.. no… is it worth addressing if it is within the capability of an advanced AAD, IMO yes.. Is the auto braking and lane departure alert capability in new cars necessary, IMO no, but there are instances where that could make the difference so I can see why they have taken advantage of the technology and implemented those features. Technology is constantly evolving, I think AADs should too where appropriate.. Divalent And here's a scenario that I wonder if your exit-algorithm would get correct: for jollies, the pilot does a zero G maneuver, so for a few seconds both plane and jumpers are falling, but in the plane. Exit detected or not? That it an excellent Question: The answer is a solid NO. Divalent And a variant (that probably is so rare its never happened nor likely to ever happen), during this zero g maneuver, a jumper sitting next to the open door floats out, but early in the maneuver, when the plane wasn't actually falling (it's upward motion was decelerating at 1g). The wind blows him towards the tail, and his head strikes the tail, knocking him unconscious. Exit detected? (So that the AAD saves his life?) In this scenario, YES, if the jumper hit the tail, they would have exited the aircraft, and an Exit would have been Confirmed. If the jumper did not slow their descent rate, the AAD would fire at the appropriate altitude. Quote Share this post Link to post Share on other sites
Fanello1775 0 #217 December 26, 2016 *** The GSA price for the various military Cypress are just over $4k (depending on specific type). Quote Share this post Link to post Share on other sites
Divalent 106 #218 December 26, 2016 Very well. Your answers imply that an "exit" is detected by means other than an altitude pattern. (Perhaps by sensing proximity to a mass of metal?) By posing those scenarios, my question really was whether a competent (or grossly incompetent!) pilot could inadvertently (or deliberately) mimic the motion an exiting jumper might have, and thus fool the device. And whether under some rare conditions an exiting jumper might mimic the motion of a plane that lost control and headed into a dive. (Although extremely unlikely, I think both are possible if the device was solely using an altitude pattern.) All things being equal, maybe it is the case that the fewer AADs on a diving plane that would fire would be better (not all reserve fires will get out the door, but maybe the guy right by the door happens to own your AAD). (OTOH, if you are in a plane going 100 mph down at 1000 ft, you are seconds away from a very bad thing happening anyway; it might even be the case that your best chance of survival would be to be ripped out of the plane at that point. But I won't take the discussion there, lol!) My only concern is that protecting against this rare situation, the additional complexity and the cost (development, testing, coding, etc; all of which ultimately will be passed along to every purchaser) might also have negative unforeseen consequences in other areas that compensates for the extremely marginal benefit it might provide. So are there other situations where determining "exit" gives an advantage/functionality over competitor's units? Quote Share this post Link to post Share on other sites
sundevil777 99 #219 December 26, 2016 QuoteVery well. Your answers imply that an "exit" is detected by means other than an altitude pattern. (Perhaps by sensing proximity to a mass of metal?) My guess, and it is just an educated guess, is that exit is detected by a combination of acceleration downward and laterally. Especially the lateral acceleration is something not easily mimicked when staying in the plane. Just a guess.People are sick and tired of being told that ordinary and decent people are fed up in this country with being sick and tired. I’m certainly not, and I’m sick and tired of being told that I am Quote Share this post Link to post Share on other sites
df8m1 24 #220 December 26, 2016 Divalent Very well. Your answers imply that an "exit" is detected by means other than an altitude pattern. (Perhaps by sensing proximity to a mass of metal?) Unfortunately I can't say Divalent By posing those scenarios, my question really was whether a competent (or grossly incompetent!) pilot could inadvertently (or deliberately) mimic the motion an exiting jumper might have, and thus fool the device. And whether under some rare conditions an exiting jumper might mimic the motion of a plane that lost control and headed into a dive. (Although extremely unlikely, I think both are possible if the device was solely using an altitude pattern.) All things being equal, maybe it is the case that the fewer AADs on a diving plane that would fire would be better (not all reserve fires will get out the door, but maybe the guy right by the door happens to own your AAD). (OTOH, if you are in a plane going 100 mph down at 1000 ft, you are seconds away from a very bad thing happening anyway; it might even be the case that your best chance of survival would be to be ripped out of the plane at that point. But I won't take the discussion there, lol!) In this sport the lesser of two evils often comes into play. Divalent My only concern is that protecting against this rare situation, the additional complexity and the cost (development, testing, coding, etc; all of which ultimately will be passed along to every purchaser) might also have negative unforeseen consequences in other areas that compensates for the extremely marginal benefit it might provide. So are there other situations where determining "exit" gives an advantage/functionality over competitor's units? Something to keep in mind our original focus was strictly on Military AAD Applications and requirements. I was “talked into” a Sport AAD at the last PIA lol. A major requirement for new military AADs is that they will not fire in the aircraft. Not to long ago there was a C-130 full of jumpers in theater (meaning they were on a real mission), and shortly after take off every AAD on board fired. The ability to identify the situation the AAD is in, is key to allowing it to confidently be a “stand alone unit”, meaning it does not have any additional or external components that either interact with the jumper, or are referenced by the AAD. It is a difference in design philosophy I would say. It would most certainly be easer and cheaper to build a standard AAD, but the military requirements have surpassed what a standard AAD can deliver. Our efforts have been focused on tailoring the AAD technology we have been working on for 10 years, to applications that can benefit from it. Applications such as Military Static Line AADs, Military Enhanced AADs, Wingsuit AADs, High Performance Canopy AADs, which all demand performance that is beyond the traditional AAD, that still to this day operate very well in the performance envelope that they were designed to operate in. It is when the jumper pushes beyond the limits of that envelope that problems start to become evident, and that is the area where our design philosophy is based. Quote Share this post Link to post Share on other sites
df8m1 24 #221 February 19, 2017 Several people asked me if I was going to keep posting in this thread because they really have enjoyed the quality of the discussion, which is a complement to everyone. I was very pleased with the feedback we received at PIA regarding the AADs we are working on. We are departing from the limitations of the 20 + year old designs that have had every drop of performance squeezed out of them. Every AAD manufacturer sent someone, (sometimes more than once lol) over to look at what we have, and some actually asked questions lol… It will be interesting to see if any of them start to adopt some of the operational features that the AADvance AAD is based on. There are some platform limitations that they have which limit what they can do, so at some point, they will need to design a new platform in order to offer most of the functionality that the AADvance AAD will. We put on two seminars, one on the SLAAD the Static Line AAD, and the other was on the AADvance AAD, which is the free fall AAD. There were problems with the scheduling that were beyond our control and I thank Travis for making the rescheduling arrangements as quickly as possible. In the end both seminars took place, and although a speaker I am not, I was told that I was able to get the information across to the audience, and I truly hope that no one left with an incorrect understanding because of how I presented the information. If anyone has any questions or would like me to clarify anything they heard, please do not hesitate to ask. We are about two months behind the schedule that I set because I looked at the designs and said to myself, this is nice, but I think it can be simpler, so I started over…. After the second design was done, there were improvements to the original design in some areas, and I went off into the weeds in others, so I took the best of the first and second designs, which is now the third design. Designing a system like this is a very challenging process. The “what if this happens?” thoughts will keep you up at night lol… This process can not be rushed, and I am glad I have the freedom to say, “ya know… lets see what this would look like if we did …” , yes it delays the product release, but when it is released, we can say that we took the time and explored every reasonable possibility, not to say that it will be the “perfect AAD”, as anyone who has build anything knows, there is always room for improvement lol.. Our new target schedule is to get the Beta units out to the test jumpers this summer and depending on how things go, a soft release could happen around a year from now. My immediate focus is on getting the ECU assembly to the programmers so they can start testing, then I can get back to working on the cutter. I do have something that I would like to ask the readers…. Although I have been trying to keep the design as compatible with the 25 year old AAD box and cutter retainer elastics, as well as the Interface window, I do not limit the design to them. The first thing that Todd (our rigger) said when he saw the head of the cutter body is “that wont fit through a cutter elastic”… and I said then change the elastic.. lol.. During the AADvance seminar a rigging question was asked and Todd answered that at one time no container was set up for an AAD and that the pockets and windows had to be installed by a master rigger. The AADvance AAD is designed to perform in all the flight modes from exit to landing on every jump. If in order to benefit from that level of performance, an elastic pocket or loop needs to be changed, then I think that is a small one time incontinence that at an earlier time was just the way it was if you wanted an AAD in your rig. I know I tend to come off arrogant when I say that, and I am curious as to what every else thinks about it. We will work with the container manufactures on any changes that “may” be required, but I am trying to keep that possibility to a minimum. I just do not think that the advancement of new technology should be dictated by the carryovers of the old. Thoughts? Quote Share this post Link to post Share on other sites
gowlerk 2,103 #222 February 19, 2017 My belief is that you need to make the cutter head backwards compatible if you want the design to be successful in the sport market. Unless your product offers significant advantages over the existing designs. There is no easy way to change the elastic keeper on containers with floor mounted cutters. The MarS cutter just barely fits the existing design. It may be possible to get manufacturers to install slightly larger keepers in new units. But only if they will also accommodate CYPRES sized cutters. One possible solution would be an included elastic that could be hand sewn into place. Or a cutter head with a regular sized appendage. But keep in mind that many containers have the cutter mounted high up and were not originally designed to be closed with a cutter between the grommets. They are already an ugly design compromise. The ideal cutter would be much flatter than the tubes currently in use. And there is no rule saying that the current elastic keeper is the only possibility.Always remember the brave children who died defending your right to bear arms. Freedom is not free. Quote Share this post Link to post Share on other sites
mxk 1 #223 February 19, 2017 I think you're making your already steep road to adoption even steeper. Unless you're using the cutter for additional functions, like having some sensors in there, I don't see how the current designs are "carryovers of the old." To put it another way, suppose Airtec and/or Advanced Aerospace Designs were given all the implementation details of your hardware and software, and were asked to produce an AAD with the same functionality. Would they also redesign their cutters and use a different elastic? This is more than just a hypothetical question, because having now seen your device, companies that are already established in this market would have either dismissed it as something that would not put a dent in their market share, or are already working on their own versions. Quote Share this post Link to post Share on other sites
df8m1 24 #224 February 20, 2017 mxkI think you're making your already steep road to adoption even steeper. Unless you're using the cutter for additional functions, like having some sensors in there, I don't see how the current designs are "carryovers of the old." To put it another way, suppose Airtec and/or Advanced Aerospace Designs were given all the implementation details of your hardware and software, and were asked to produce an AAD with the same functionality. Would they also redesign their cutters and use a different elastic? This is more than just a hypothetical question, because having now seen your device, companies that are already established in this market would have either dismissed it as something that would not put a dent in their market share, or are already working on their own versions. Attached are two pics. The cutter on the top is an Argus cutter, the center one is my initial concept cutter, and the bottom one is Airtec's. Gowlerk and MXK, The cutter design has no relation to the AAD operational functions; it is strictly what I would like to see in a cutter. I will put it back on the lathe and turn down the corners a bit more and see what it looks like. The reason for the flat on top is that, 1, it is easy to machine a nice radius around the hole, and 2, I do not like how a hole through a cylinder creates a saddle of sorts that a grommet can be pushed down into and wedging the loop against the cutter body. I imagine that the flat also will tend to keep the cutter’s projected loophole centerline parallel with the closing loop if that makes any sense lol… This cutter has a larger loop hole than the Airtec cutter as well as a larger blade piston (single edge V). I need some wall thickness at the end for a blade locating pin that will keep the blade in place until the cutter is fired. An alternative practice to locate the blade is to pinch it by compressing dimples into the sides of the cutter body and I do not like that idea. There is no way to test the amount of pinch. I like a shear pin much better. I will take some off of the corners and see what that looks like, The main stem diameter is the same as the Airtec brass section so that is not problem, it is the corners that cause an issue. One work around would be to feed the cable through the elastic and piping connector first. There could be some kind of pointed cap that screws on the connector and a pull up cord could be pushed through the piping with a gun cleaner and then the connector pulled through?.. Noting is set in stone and if I can retain the features I want and improve compatibility with existing containers then that is a no brainier.. And if Airtec or anyone else for that matter wants our secret recipe we are open to selling it. This has all been funded by private investment and if it makes business sense to sell out to the competition, the so be it. The US government wants to own the next generation of AADs that are fielded so if we win the Static Line AAD contract, we will be selling them rights to the design as well. Same goes for the free fall AAD, the Gov wants to be able to control the quality and get competitive bids on building them.. How that will work out in the end is anyone’s guess but that is the plan. There is no doubt that the others are looking at how they can implement what we are doing, and that is a win for everyone really. There are some aspects that they cannot do with the hardware they currently have. And depending on how they determine the actual altitude of the jumper in free fall, it might be more of an effort than they really want to put into a design that is in realty close to the end of its run IMO. I am not sure if it is ok to attach the power point presentations to this thread or not. I imagine this is walking a fine line between advertising and a technical discussion as it is lol… Quote Share this post Link to post Share on other sites
pchapman 278 #225 February 20, 2017 For everyone comparing cutters, we clearly need to whip out the M2's flat sided cutter as well. A pic from the web is attached. (From a review of the M2 at skydivenow.co.uk ) Quote Share this post Link to post Share on other sites