AR Gun Scope

Overview

Background

While leading the prototyping department at Third Eye Gen, I was tasked with converting hardware from our X2 AR glasses into a gun scope prototype for the first and second phases of a defense contract. The goal of this project was to fabricate a scope that could detect vehicles and overlay sights using an augmented reality display, simulating the experience of firing a rocket launcher for training purposes.

Results

In just 10 days and a $1,500 budget, my team successfully delivered phase 1 and 2 prototypes and won a $1.5M contract.

Figure 1: Concept render of rubberized aluminum scope body

Figure 2: Phase II fully functional prototype delivered to customer

My Role

With only 10 days to fabricate and ship a functional prototype, I shared the role of lead engineer with a co-lead with a background in mechanical engineering. We worked alongside each other splitting the work of CAD design, electronics mounting, thermal simulations, and fabrication to iterate on multiple parts at a time. I utilized the in-house 3D printer mini-fab I had previously set up to rapidly test and iterate through designs.

The 10-Day Sprint to Valhala:

Day 1 - The Potato

Figure 3: The Potato - A day 1 prototype to pass off to the software team for integration testing.

After commandeering the electronics from several AR headsets, it was clear that component placement would be severely constrained by the unmodifiable cable length connecting each component. Most connections were extremely short and delicate and couldn't be altered without a lengthy redesign.

Day 2 - Modular Test Bench

Figure 4: A modular testing rig to figure out component placement.

Each 3D printed scope took 12+ hours to print, so I made a modular testing rig to quickly iterate on component placement without having to wait on the printer.

Day 3 - Electronics CAD and Mounting

Figure 5: Preliminary cad placement of internal components

Honing in on hardware mounting, I modeled the placement of each component after considering cable length, thermals, assembly, and durability. This layout drove the rest of the scope's design.

Day 4 + 5 - Minimum Viable Product (MVP)

We're halfway there with one fully functional 3D printed prototype that gets hot enough to melt its plastic housing if left turned on and unattended. It's not pretty and was an absolute pain to assemble, but the software team was excited to get an upgrade from The Potato.

5 days left to fix thermals, add a control interface, and get the entire scope body milled from aluminum.

Figure 6: Iterative design graveyard

Figure 7: An iteration showing layout is improving

Figure 8: First fully-functional 3D printed prototype

Day 6 - Thermal Managment

Before removing the AR compute module from its original housing designed to dissipate heat, the processor already had issues staying cool. Taking that processor out of its natural habitat and slapping it into an enclosure right next to a bunch of other thermally sensitive components went just as well as one could expect.
The AR compute module was originally designed to be passively air-cooled through natural convection. This made sense for an ultralight plastic AR headset but wasn't a requirement for our scope.
I designed a custom heatsink to interface with the AR processor's existing thermal package and provided my co-lead with each component's approximate wattage to validate sufficient heat dissipation.
No active cooling was required because the gunscope's (future) aluminum body had enough thermal capacity to dissipate the processor's heat without getting much warmer than room temperature.
Since we only had one shot with the aluminum body, I added a port for a small cooling fan so it could be added later if needed.

Figure 9: Custom heat sink to replace convection cooling with conduction cooling through the scope's aluminum body

Figure 10: Steady-state thermal distribution

Figure 11: Simulation modeling air convection

Figure 12: Simulation showing effects of ventilation holes and a low power fan

Day 7 - Design Optimizations and CNC Milling

I modified the design of the final scope body to be manufacturable on a 4-axis CNC mill to meet budget requirements.
With 3 days left before the final scope had to ship, I worked with a local machine shop, Draftek Designs, to mitigate potential shipping delays. Their personalized help and expedited accommodations were paramount to this project's success. Thank you Draftek!
Coming from an electronics background, climbing a steep learning curve in a new discipline was incredibly rewarding. The final scope design was submitted for fabrication on day 7 and would be ready for pickup the morning of day 10.
This step was probably the hardest part of the entire project. There was no room for mistakes and the scope body is one large complex piece.

Figure 13: Final CAD design of the AR gunscope optimized and sent off for CNC fabrication

Day 8 + 9 - Control Interface and Silicon Molds

While the aluminum body was getting milled, I was able to focus on improving user experience.
I added a square hole with a lip and screw holes in the machinist's file to give myself flexibility with the design of the button interface. This was also a great place to mount the 2.4Ghz WIFI antenna so the signal could penetrate the aluminum enclosure.
Using 3D printed molds, I was able to make a silicon eye-piece in-house to protect the user from the metal scope and eliminate light bleed when viewing the AR display.

Figure 14: Button and charging interface

Figure 15: 3D Printed Molds for Eyepiece

Figure 16: Silicon Eye-piece from two-part silicon resin

Day 10 - Assembly and Final Testing

Figure 17: The aluminum scope body fresh from the machine shop

Figure 18: Putting it all together

Figure 19: Finished prototype

Page updated

Google Sites

Report abuse