Howdy 👋 

It ain’t “Tech Tuesday” but I hope you’ll enjoy this blast on your Wednesday evening or Thursday morning.

I’m Digital Dave back from another writing hiatus. I’ve needed some time to organize my thoughts. A bunch of ideas have been bouncing around in my brain…almost as many balls I’ve been bouncing around the bocce courts.

I also am fresh back from the Las Vegas National Bocce Tournament where my team (Silly Sticky Guys) placed 5th in Silver. We played a whopping 14 games across 4 days. I think our record was 8 wins 6 losses. Scroll for pics.

Welcome to the 4 new subscribers since the last newsletter blast.

Last time I shared some goals for the project:

While I joked about setting ambiguous goals:

This time I’d like to set some ambitious goals.

System Status

On Sunday in Las Vegas, I had the opportunity to speak with Michael live on the Bocce Broadcast Network about bocce and about Bocce Labs.

One of Michael’s questions stands out: “So how far are we from seeing this as a reality?”

I stumbled a bit (go watch & listen for yourself). So here’s a more organized response crafted just for Sharkey.

The truth is, that you can see the reality of what I’ve built in my living room. Just invite yourself over for dinner and we can geek out.

The status of what I’ve built is as follows:

• The system finds balls reliably

• The system measures distances between balls reliably with precision to less than ¼ inch.

• The system can distinguish between balls that are different enough in colors (plural colors…patterns are fine and encouraged, but if the patterns have similar colors, then we have an issue)

• I could train AI to recognize ball patterns agnostic of colors, but the algo wouldn’t be super fast, so I haven’t spent time on that

• I don’t have the interface to a digital scoreboard, but I do have a working digital scoreboard (one for bocce and one for curling)

• The FPS (Frames Per Second) is not fast enough for raffas (i.e., fast moving balls), but various algos can be turned off so that you can at least see the fast moving ball

• The system can interface with a broadcast system. Both OBS and Switcher Studio are supported (neither have been tested).

And that leads to my next thought.

Lab to [Minimum Viable] Product

What if I don’t need to wait for all kinks to work out?

What if the best option is to release what I have now into the world so that I’m motivated to take things to the next level (and find collaborators to chip in)?

I’d like to make that happen. Just keep in mind that I’m not a Product Manager. I’m an Engineer.

That said, if I put myself in the shoes of a Product Manager, here’s what I’d tell my engineers they need to do.

It was helpful to sit in the broadcast booth in Las Vegas and comment on the majority of the first silver finals game. I have an idea what a broadcaster needs.

I’ve been an ABC Referee since early 2021. I don’t referee as much as I did in 2021 anymore, but I know what would help a referee.

I also have an idea of what I’d want as a player and/or league/tournament manager. Umm, hello, knock knock stats and a website to view them.

So with that, I’m presenting a Minimum Viable Product (MVP) roadmap. In fact I have an upcoming visit to Waterside Club near Syracuse, NY to consider some of these steps.

MVP 1

Camera & Software Assisted Referee Measuring Aid (iPad/tablet based) - A referee makes every decision but can look at the screen & computer measurements to help make those decisions. The referee knows the precision & tolerance of the system.

Broadcast Integration - For OBS, that’s a simple HDMI out from the computer to HDMI in to the broadcast switcher and switch to the feed as necessary. For Switcher Studio, just install the desktop plugin and switch to the feed as necessary.

AI Data collector - Collects ball image data as needed and stores for AI training.

MVP 2

Everything above.

Ball Distinguisher - I’d improve the ball distinguisher because I’d have data from the system in more scenes and lighting scenarios (not to mention more ball patterns).

MVP 3

Everything above.

Integrate with TV Digital Scoreboard - Like I said above, I have a TV Digital Scoreboard software that I’ve written. It includes a wireless remote & receiver. Ideally, I’d make it so the ref could control the scoreboard from the iPad/tablet too.

MVP 4

Everything above.

Score an entire game automatically (with Referee/Player Override) - I think I’ve been too focused in this area. As you can see there are three Minimum Viable Products that come before this.

MVP 5

Everything above.

Tolerance to 1/8th of an inch - Currently the system is precise to ¼ inch but flickers of light cause the numbers to jitter. Additionally, a subscriber named Richard explained how the coordinate system needs to be mapped. I tried to explain on the BBN broadcast that tolerance directly under a camera is slightly different than 6 feet to the side. It isn’t a simple proportion of the camera pixel coordinates to the bocce court coordinates. So the coordinate system needs to be mapped with math.

MVP 6

Everything above.

Game data logged to a Cloud Database - This just involves a REST API to get the data from the court up into the cloud. It involves some database design (most of which I’ve thought through and even experimented with in 2021).

Stats website where players can access 3 stats - Create a React website for players to access their stats after a game. An even simpler approach to this would be a live spreadsheet as Python easily interfaces with Google Sheets.

Version 1.0

Everything above.

Sell the product and service to bocce clubs. Charge for the installation and service contract. Have a plan such that there’s a path to revenue to paying for ongoing development, maintenance. This is the hardest part. My friends at AccuTennis have a killer engineered system that provides a lot of value, but the business model is the hardest part. Not a lot of tennis clubs are able to afford AccuTennis.

Next Steps

My next step is to go Waterside Club in New York to explore the feasibility of installing cameras there and how we might get some university engineering students involved in the project. Bocce Broadcast Network is eager to connect my system as a broadcast aid and I’m confident that I can deliver MVP 1, 2, and 3 in a reasonable amount of time.

That’s Vegas Baby

See you soon

Thanks for reading! Please share this newsletter with a friend. Just smash that FWD button and drop it in an email or share your referral link above.

CLICK » Watch me play and Listen to me chat about Bocce Labs

~ Digital Dave