If you spend any time on golf TikTok, you may have come across Coleman Rollins documenting one of the more ambitious DIY projects in the space: building a home golf launch monitor completely from scratch.

With regular real-time video updates shared with his 11,000+ followers, Coleman has built a dedicated audience to see just how far you can push a garage build when curiosity, persistence, and a bit of engineering obsession collide.

I caught up with Coleman for a quick Q&A to dig into the realities behind the project: the time invested, the true cost, the technical challenges, and the advice he’d give to GolfWRXers tempted to try something similar themselves.

@colemangolfs (Reupload) Part 16 – building a launch monitor. Did we just do this??? #diy #golfgear #diyproject #golftok #golf @Foresight Sports @Bushnell Golf @Garmin Official ? original sound – Coleman Rollins

• Follow Coleman’s journey on TikTok here, and Instagram here.

Gianni: What made you decide to try and build your own home simulator instead of buying one?

Coleman: I was going through a cycle every 3-5 months of wanting to buy a launch monitor. Every time I went to look back over the options, every single one had a downside. Either too expensive, requires a subscription, or doesn’t have the metrics I wanted. Not only that, but they all have proprietary software that is often pretty bad. With my background as a software engineer, I decided to just give it a shot and see what happens.

Gianni: How daunting was it at the start, figuring out all the parts you’d actually need?

Coleman: It was really daunting at first, since I knew nothing about hardware projects or radars! But thanks to the power of AI and LLMs, I was able to get a jumpstart on what I might need. It became pretty clear right away that the most cost-efficient and quick-to-start option was a radar-based device (rather than a camera-based).

Gianni: How many total hours are you into this build so far, including the parts that didn’t make the final cut?

Coleman: This is a tough one to estimate, but probably at least 150 hours. I had some PTO left at the end of last year and spent a considerable amount of time every day in December on it.

Gianni: What’s the real all-in cost right now, and how far off is that from your original budget?

Coleman: The all-in cost for the parts currently in use is around $350. I’m really happy with this number. Originally, I wanted to do it for around $200, but I decided early on to go with a nicer radar that cost a bit more, and it has been worth it. I would say another $200 – $300 has been spent on parts that aren’t used, various software-related services, and training machine learning models.

In Part 38, Coleman improved the ball tracking capture from 30 to almost 180 frames per second.

Gianni: What sensing method are you using at the core of the launch monitor, and what trade-offs did that choice force you to accept?

Coleman: The core method is Doppler radar. There’s a lot of debate (including in my comment section) about whether or not the camera is better than the Doppler. Trackman devices are proof that Doppler is a viable high-end option, despite the majority of other high-end launch monitors being camera-based. The biggest tradeoff I have found is that camera-based methods need much more computing power and higher-end hardware. You need very high frame rates to track the ball, and a lot of computing power to process each frame quickly. For one guy on a budget, Doppler was the clear choice.

Gianni: What part of the system is currently limiting the overall accuracy the most?

Coleman: Right now, I’m spending most of my time dialing in the radar. Something cool I have learned is that each moving object has a unique radar frequency. If you make a bunch of swings with a golf club in front of the radar while recording the frequencies you see, you can start to see a pattern in what a swing looks like. Then our job is to get our software to recognize that and process the data.

Gianni: Where do you think most people trying to copy this build would get it wrong?

Coleman: There can be small nuances between various pieces of hardware or small differences in radar models, so that’s something to look out for. But I’ve spent a lot of time documenting everything that people need, steps to get started, and where the code lives, so I hope it will be easy to follow!

Gianni: What’s the biggest piece of advice you’d give to a GolfWRX member inspired to try this themselves?

Coleman: Get in there and go for it! It may require some patience at times, but I have learned so much in the three months I’ve been at it, which has been amazing. All the steps I have taken and the code for the device are all free, so if it’s something that sounds fun and interesting, the tools are there.

@colemangolfs Part 40 – building a launch monitor. Will this be our final form? ??#golf #golftok #golftech #claude ? original sound – Coleman Rollins