My Experiment with Building a Coin Sorting Machine by Date

As a coin collector who enjoys tinkering, I recently decided to try building a machine that sorts coins by date and mint mark. What began as a fun side project combining numismatics and electronics turned into a fascinating challenge. Let me walk you through what I discovered, including some tips that might help in your own collecting adventures.

The Technical Challenges I Faced

Getting the mechanics right was trickier than I expected. The real headache? Creating an image recognition system that actually works. You’d need a massive library of coin photos to account for differences in wear, lighting angles, and positioning. Here’s what became clear:

• Each coin side requires at least 360 images to cover full rotation, plus separate sets for different wear levels—like fresh mint state versus pocket-worn coins
• Reading dates and mint marks demands sharp edge detection and OCR tech, but the necessary high-speed cameras alone can cost more than a nice Morgan dollar collection
• I’m testing a three-bin approach: definite matches go to ‘accept’, questionable ones to ‘unknown’, and clear mismatches to ‘reject’—just like sorting wheat pennies by hand

It struck me how this could potentially spot VAMs or minor errors on Morgans automatically—like having a tireless assistant searching for varieties!

Financial Realities and Market Potential

When I calculated costs versus benefits, the numbers didn’t look promising. Parts alone would run several thousand dollars, and sourcing sufficient coins is challenging—banks like Brinks just want rolled coins, not date-sorted ones. My honest assessment:

• For personal use, you’d spend years hunting rolls just to break even after equipment and transportation costs
• While some collectors might pay $5,000+ for advanced features like VAM detection, it’s a tiny market. Software licensing could be smarter than hardware sales
• Treat this like any passion project—similar to building model trains—where the satisfaction comes from creating, not profiting

The whole experience reminded me why we collect coins: for the history and joy, not get-rich-quick schemes.

Practical Advice for Fellow Collectors

If you’re considering automated sorting or roll hunting, here’s what I’d suggest based on my trial-and-error:

• Master manual sorting first—know your key dates and mint marks before automating. Hunting pre-1982 copper cents? A machine probably costs more than you’ll gain
• Machines struggle with wear patterns. Always inspect valuable coins yourself—your eyes beat any camera for grading accuracy
• Start small with error hunting. A $10 loupe or free coin app works better than jumping into complex machinery

One more thought: if these machines become common, sorted coins might flood the market. Sometimes the old-fashioned approach preserves both value and the thrill of discovery.

Where I’m Heading Next

Despite the hurdles, I’m still tinkering with a prototype. I’ll start simple—letting users search for specific dates or mints within one denomination. Small test batches will help refine the system. If it works, it could be handy for fellow collectors. But honestly? I’m doing it for that electric moment when the first coin clinks into the right bin—the same excitement we feel flipping through a fresh roll of pennies. The journey’s the reward.