A new payload interface design was required to make use of Astro's powerful flight controller and internal Linux computer. To make the most of this brand new interface, an ambitious set of requirements was developed from the project onset: 

• 40 electrical signals needed to run through the connector and the interface has a lifecycle over 2500 connections.
• It required a safety that would automatically engage when the payload was inserted in addition to a second action that would fully secure the interface.
• The entire design needed to be simple enough to be made into a standard that 3rd party drone or payload manufacturers could successfully interface with or fully recreate. 
• The user needed to be able to operate the interface with a single gloved hand.

To make sure that the payload interface could accomplish all of these requirements I designed a sliding dovetail quick release with a spring loaded safety that would engage at the point where two halves of the connector made full electrical contact. The whole interface would be secured using an over-center clamp. Due to the nature of such an interface, it needed to account for a large tolerance stackup to ensure the connector did not bind or damage itself during installation or when the overcenter clamp was engaged. To account for this stackup, I chose a connector that had 0.5mm of float in the X and Y axes and added a molded part to the system to increase the compliance by 0.5mm in the X, Y, and Z axes. The molded part was designed to deflect first and then minimize the forces applied to the connector itself, thus increasing its lifetime. 

Once the interface was designed and tested it was added to the Pixhawk Payload Bus Standard as its official mechanical interface standard. As of the start of 2024 this interface is being used by four other payload manufacturers to mount and control lidar, SLAM payloads, and a variety of gimbals as well as another drone manufacturer for its payload interface.

Much of Astro is made from 6061 aluminum as it is easy to machine, lightweight, and relatively inexpensive. However this choice of material came with a drawback: that part fatigue is not a question of ‘if’ but rather 'when' . To ensure Astro would not have any part fatigue during its lifetime, I identified the parts and assemblies that would require detailed analysis and tests; these included the upper/lower chassis plates, prop adapters, and boom hinges.

I used various methods to ensure that my parts would meet Astro's 5 year lifetime warranty with an additional safety factor of 1.5 to avoid failure events for our users. The most efficient check performed was a static load calculation for the prop adapters and the boom hinge assemblies; this would highlight any obvious issues with the design and give us an indication the parts would stand up to continuous flight loading. Once acceptable static load stress was reached, I captured the vibration data from an out of balance propeller during an aggressive flight and set up a test where I used offset weights and motor clocking to recreate the worst case instances captured during the data logging flight. This test was run for over 100hrs until the number of cycles in this worst case state met our estimated cycle count from an aircraft that was being flown in an extreme manner for 7.5yrs of flights. 

Before starting to empirically test the prop adapters, I ran an additional set of calculations to determine the dissymmetry of lift, which is an important part of the the forces that act on the prop adapter. This dissymmetry is a result of the difference in relative airspeed that the advancing and retreating prop blades experience, and ends up imparting a high count cyclical load to the prop adapter. Once I calculated these forces for Astro's top speed, I was able to run Finite Element Analysis (FEA) to ensure that the prop adapter design would stand up to the empirical validation test and flight hours we would subject it to.