To our customers,
We endeavor to be honest with our product schedules and aircraft autopilot expectations, and have updated the upcoming autopilot approvals to indicate when things change. But, we could have better set expectations, and better explained why dates have slipped and why airplane models have fallen off the list of expected approvals. For that we apologize, and we’re working to improve.
There are fundamentally two things that we’ve tried to give guidance on: airframe compatibility and availability timeframes. Both have had their own reasons for being too optimistic in the past. We’ll explain why our estimates haven’t been right, and how our approach is changing now and in the future.
On timelines
To arrive at an estimated ship date for an autopilot approval project, we generally take our internal schedules, seriously pad them, round forward to the next quarter/half/year as we estimate based on risk, and publish that as our best guess. A few things have conspired to make these too optimistic.
First, although COVID didn’t affect our own work drastically, it did substantially slow and shut the FAA’s operations for a while. Another factor here was the aircraft grounding issues experienced by our neighborhood airline manufacturer, which is serviced by the same Seattle certification office that handles our projects. As a result, the staffing and attention available to our projects has been less than we anticipated. We based our timelines on the results of our original 172 and Bonanza 35 programs, which moved fairly quickly, and that turned out to be the wrong way to calibrate the future. We are working with the FAA to address this going forward.
Next, there were a couple of projects that we expected to be able to handle with external resources and partners, which we later realized we’d need to bring inside Dynon to perform the deep design and engineering analysis required. This turned a few parallel projects into ones that had to be serialized, which impacted the schedules of all projects. This is also why we are currently unable to entertain offers to partner with external parties to try to speed up aircraft approvals. So far, we’ve found it counterproductive. The level of specialized expertise is too deep for most individuals and organizations that don’t have a lot of experience in this exact space.
On aircraft models
We’ve aimed to be as transparent as possible in terms of what aircraft we thought we would be gaining approval for with each effort. Our ideal scenario would let potential customers plan their whole installation knowing whether or not an approval was coming. We recognize that for many people, the autopilot is important in the solution, and so we want to help you understand which airplanes will have those capabilities as early as possible. For a while, this was the way we communicated our plans. But we overshot. We were naive in thinking that our initial analysis for a given model family represented where we’d end up. Along the way, we’ve found unexpected changes across aircraft sub-models/years as we’ve gotten into the deep engineering. We’ve also discovered differences across serial number ranges of single models that should be identical, but which turned out to have differences that matter for the autopilot install. We found we are dealing with legacy aircraft that were essentially handbuilt with variations from plane to plane. These findings have led us to reign in aircraft model expectations. With 20/20 hindsight, we should have been way more conservative up front. We will be in the future.
What the future looks like
We’ve learned a lot over the past couple of years, and things are looking better going forward. We expect our cadence to improve with less project shuffling, increased staffing, applying lessons learned, and leveraging the community as best we can. More on that below.
Know that if an airplane is on our website, it’s one we’re actively working on, we have a version of the airplane in-house being worked on. Going forward, we will make guesses on what related/similar models are likely to be included, because we’ve had some late dropouts as happened with the Bonanza series. We’ll publish known exclusions so that we don’t offer false hope when we know certain models or variants won’t make the approval. We want you to have that information early. Conversely, just because a model isn’t excluded, doesn’t mean it’s “in”. We’ve updated our language on the website to set these expectations better.
Whenever we know that a schedule has changed, we’ll update our website like we always have. We’re considering removing date-based guidance entirely since we’ve been so consistently wrong historically. At the same time, we want to provide the best guidance we have, even if it’s low confidence. We’re curious what you think we should do here?
We are also considering not publishing future aircraft models before they’re available. We’re curious what you - the community - think of that idea. On one hand, you won’t know whether or when your aircraft will have autopilot coming. On the other, if autopilot is important to you, it’s not REALLY available until it’s certified for your model. And we can’t be certain of which models we’ll gain approval for until pretty late in the process. Your thoughts?
Finally, know that ever since the Bonanza autopilot applicability narrowed, our team has been actively discouraging potential customers from buying in advance of actual autopilot availability if an autopilot is important to their project. We’ve strengthened the language on the website similarly. But we know that some of your panels predate that shift in expectations. Some Bonanza 33, 35 and other customers have bought and installed our products, expecting an autopilot that hasn’t yet been approved. We’re really sorry, and we are working to deliver on our original plans. What does that mean, specifically? We currently have a pretty full roster of airplanes, in hangars, being worked on, in various states (the ones on our website). We’re committed to all of these airplanes. As those finish up, we plan to come back to as many of the aircraft that we previously believed were in our certification projects, including as many of the the Bonanza 33 and 35 series as we can. As of this writing, we do not have the precise order completely mapped out, but we will update the community when we do. It is also possible that some aircraft that we’d previously thought we’d gain approval for will not ultimately gain it, since there may be some small populations of aircraft that are too different from the rest of the fleet of that model.
Help us understand your aircraft
Many of you have asked how you can help. Some have generously offered to loan us your aircraft, provide services, or otherwise partner with us. We appreciate all of your passion and offers. Some of those are impractical from a certification standpoint, but there is at least one meaningful way you can help. One of the “gotcha” areas is seemingly minor differences in airframes across model years that can greatly affect autopilot design and applicability for the whole family. Oftentimes, robust drawings and structural data for entire model lines are hard to obtain. Having a better look inside the “guts” of as many aircraft as possible would help a lot here. This will hopefully let us expand and verify the exact models that future autopilot approval projects will apply to.
So what do we need? Pictures of the inside of your fuselages, wings, and the other structures that autopilot servo installations live in. The best pictures would be those of existing autopilot installations. If you can include a ruler for scale, that would be helpful too. Finally, knowing the actual diameters of your aircraft’s control cables would make our engineers really happy. The best time to grab such pictures is probably during your annual or other significant work. If you’re game to help, the image uploader can be found here. Thank you in advance!
We endeavor to be honest with our product schedules and aircraft autopilot expectations, and have updated the upcoming autopilot approvals to indicate when things change. But, we could have better set expectations, and better explained why dates have slipped and why airplane models have fallen off the list of expected approvals. For that we apologize, and we’re working to improve.
There are fundamentally two things that we’ve tried to give guidance on: airframe compatibility and availability timeframes. Both have had their own reasons for being too optimistic in the past. We’ll explain why our estimates haven’t been right, and how our approach is changing now and in the future.
On timelines
To arrive at an estimated ship date for an autopilot approval project, we generally take our internal schedules, seriously pad them, round forward to the next quarter/half/year as we estimate based on risk, and publish that as our best guess. A few things have conspired to make these too optimistic.
First, although COVID didn’t affect our own work drastically, it did substantially slow and shut the FAA’s operations for a while. Another factor here was the aircraft grounding issues experienced by our neighborhood airline manufacturer, which is serviced by the same Seattle certification office that handles our projects. As a result, the staffing and attention available to our projects has been less than we anticipated. We based our timelines on the results of our original 172 and Bonanza 35 programs, which moved fairly quickly, and that turned out to be the wrong way to calibrate the future. We are working with the FAA to address this going forward.
Next, there were a couple of projects that we expected to be able to handle with external resources and partners, which we later realized we’d need to bring inside Dynon to perform the deep design and engineering analysis required. This turned a few parallel projects into ones that had to be serialized, which impacted the schedules of all projects. This is also why we are currently unable to entertain offers to partner with external parties to try to speed up aircraft approvals. So far, we’ve found it counterproductive. The level of specialized expertise is too deep for most individuals and organizations that don’t have a lot of experience in this exact space.
On aircraft models
We’ve aimed to be as transparent as possible in terms of what aircraft we thought we would be gaining approval for with each effort. Our ideal scenario would let potential customers plan their whole installation knowing whether or not an approval was coming. We recognize that for many people, the autopilot is important in the solution, and so we want to help you understand which airplanes will have those capabilities as early as possible. For a while, this was the way we communicated our plans. But we overshot. We were naive in thinking that our initial analysis for a given model family represented where we’d end up. Along the way, we’ve found unexpected changes across aircraft sub-models/years as we’ve gotten into the deep engineering. We’ve also discovered differences across serial number ranges of single models that should be identical, but which turned out to have differences that matter for the autopilot install. We found we are dealing with legacy aircraft that were essentially handbuilt with variations from plane to plane. These findings have led us to reign in aircraft model expectations. With 20/20 hindsight, we should have been way more conservative up front. We will be in the future.
What the future looks like
We’ve learned a lot over the past couple of years, and things are looking better going forward. We expect our cadence to improve with less project shuffling, increased staffing, applying lessons learned, and leveraging the community as best we can. More on that below.
Know that if an airplane is on our website, it’s one we’re actively working on, we have a version of the airplane in-house being worked on. Going forward, we will make guesses on what related/similar models are likely to be included, because we’ve had some late dropouts as happened with the Bonanza series. We’ll publish known exclusions so that we don’t offer false hope when we know certain models or variants won’t make the approval. We want you to have that information early. Conversely, just because a model isn’t excluded, doesn’t mean it’s “in”. We’ve updated our language on the website to set these expectations better.
Whenever we know that a schedule has changed, we’ll update our website like we always have. We’re considering removing date-based guidance entirely since we’ve been so consistently wrong historically. At the same time, we want to provide the best guidance we have, even if it’s low confidence. We’re curious what you think we should do here?
We are also considering not publishing future aircraft models before they’re available. We’re curious what you - the community - think of that idea. On one hand, you won’t know whether or when your aircraft will have autopilot coming. On the other, if autopilot is important to you, it’s not REALLY available until it’s certified for your model. And we can’t be certain of which models we’ll gain approval for until pretty late in the process. Your thoughts?
Finally, know that ever since the Bonanza autopilot applicability narrowed, our team has been actively discouraging potential customers from buying in advance of actual autopilot availability if an autopilot is important to their project. We’ve strengthened the language on the website similarly. But we know that some of your panels predate that shift in expectations. Some Bonanza 33, 35 and other customers have bought and installed our products, expecting an autopilot that hasn’t yet been approved. We’re really sorry, and we are working to deliver on our original plans. What does that mean, specifically? We currently have a pretty full roster of airplanes, in hangars, being worked on, in various states (the ones on our website). We’re committed to all of these airplanes. As those finish up, we plan to come back to as many of the aircraft that we previously believed were in our certification projects, including as many of the the Bonanza 33 and 35 series as we can. As of this writing, we do not have the precise order completely mapped out, but we will update the community when we do. It is also possible that some aircraft that we’d previously thought we’d gain approval for will not ultimately gain it, since there may be some small populations of aircraft that are too different from the rest of the fleet of that model.
Help us understand your aircraft
Many of you have asked how you can help. Some have generously offered to loan us your aircraft, provide services, or otherwise partner with us. We appreciate all of your passion and offers. Some of those are impractical from a certification standpoint, but there is at least one meaningful way you can help. One of the “gotcha” areas is seemingly minor differences in airframes across model years that can greatly affect autopilot design and applicability for the whole family. Oftentimes, robust drawings and structural data for entire model lines are hard to obtain. Having a better look inside the “guts” of as many aircraft as possible would help a lot here. This will hopefully let us expand and verify the exact models that future autopilot approval projects will apply to.
So what do we need? Pictures of the inside of your fuselages, wings, and the other structures that autopilot servo installations live in. The best pictures would be those of existing autopilot installations. If you can include a ruler for scale, that would be helpful too. Finally, knowing the actual diameters of your aircraft’s control cables would make our engineers really happy. The best time to grab such pictures is probably during your annual or other significant work. If you’re game to help, the image uploader can be found here. Thank you in advance!
