Testing Six Hundred Fish

fish

That’s the best and most obtuse title you’ll ever see for a Hackaday post, but surprisingly it’s pretty accurate. [Bob] over at the Sector67 hackerspace took part in a 111-day accelerator program in Shenzhen last year to improve his manufacturing skills. He’s just about ready to release his first product, a Bluetooth device that connects to an ice fishing tip-up. The blog for the device recounts the challenges of taking a project from a circuit to a finished product, and illustrates the difference between building something with an Arduino and selling thousands of devices.

The circuit boards for BlueTipz come in panels of eight, but what’s the best way to populate and solder five thousand devices in a reasonably short amount of time? [Bob] hired a few students from the local college to help him out in assembling all these devices. The plastic enclosures were made at a local plastics manufacturer, but the molds were made in China. The manufacturer needed to modify the molds a bit, but after a few days, [Bob] had five thousand enclosures ready to stuff full of electronics.

With the devices assembled, it’s time for programming, and that means building a programmer. [Bob] put all the guts for the device into a plastic box and 3D-printed a mount for the bare BlueTipz board. Put a board on the mount, press a button, and the tech now has a functioning device in his hands.

Besides manufacturing, there’s also a whole lot of testing that went into the design of BlueTipz. Because this is a device for ice fishing—a cold and potentially windy operating environment—[Bob] built a test rig in a freezer. The test rig triggers the device’s sensor, waits two minutes (the amount of time it would take for an ice fisherman to check the tip-up) and resets. They claim the battery life is good for 600 fish, and with this testing rig they were able to verify their calculated battery life with real-world data: without actually catching six hundred fish, of course.

Not only does [Bob] have a good bit of product development under his belt, he was also kind enough to go over the stuff everyday electronic design just doesn’t cover. Developing a product is something you can only learn by doing, and we’re glad [Bob] chose to share his experiences with us.

16 thoughts on “Testing Six Hundred Fish

    1. Hirudinea, I think it _enables_ that culture even more. Now they can continue to drink and not have to worry so much about checking their tip-ups every couple minutes, secure in the knowledge that their phone will tell them when they need to do something.

  1. Nice write up. Although it’s probably TMI to discuss, I wonder what the manufacturing/assembly/testing/packing cost was compared to their retail selling point of $40 per unit.

    1. I can’t go into too much detail, but we approached the price point from two sides. We looked at what the closest competition was selling and priced ours near theirs but higher because we had more features and better performance and newer technology. Then we talked to retailers and figured out where they wanted to sell (which turned out to be higher than we expected, but we’re not complaining about that), then what they needed to buy at wholesale price from us. So between retailers, distributors, and sales reps all taking a cut, and our cost of goods sold, the profit isn’t as great as you’d think. We have significant tooling and inventory costs that we need to recover as well, so our first year certainly won’t be making us rich.

  2. Looking forward to reading about your “adventurers” into the world of manufacturing. Find ourselves in a similiar position of having to go from a prototyped proof of concept design to a unit that can be built and delivered in a timely and reliable manner. Would be nice if this operation could run at a profit, as I have a lifetime of experience running at a loss (2nd law of thermodynamics?) and could use a change….

    Seriously, there is a huge difference between between hobby hacking, and making a profit with OSHW. If I am paying attention, I learn something new every day. I hope your article will provide that learning this morning.

    Good luck.

        1. Most are coming from stores. We are the only major online retailer (by design), but we are doing healthy online sales, and we can sell 2 in stores for every 1 we sell online and make the same profit (we can’t sell online cheaper than stores, though, because the stores hate being undercut). Some of the smaller stores have come to us after customers have asked them about our products (those are the best). For some we have sales reps that helped get us in. For others it was a distributor. With the big chains, they decide their purchase orders for the ice fishing season by March of the year before, so we had no chance of getting in some of them this season, and we are working on getting into them next season. The biggest thing is to make connections with people and work your way up the chain. Going straight to the top doesn’t usually work.

  3. Excellent write up, thanks!

    One question is rarely answered in these type of posts, which often start with “we found a supplier….” or “we contacted a manufacturer”, and it’s surely one of the first, and most important steps in the journey.

    How do you go about finding a manufacturer or supplier? Word of mouth? Yellow pages? Search on Alibaba? And what should you be asking them?

    And how do you approach parts you’re not even sure how to manufacture? For example, I have a project idea that will need some optics. I don’t even know how to design the thing I need, let alone how to manufacture it, or a manufacturer to contact.

    1. We have a lot of suppliers for our simple product, and used a lot of methods. For the rolling ball sensors and the LEDs, we got samples of every kind we wanted, tested them, and went with the one we wanted. There were some long Skype chats with a Chinese rolling ball sensor manufacturer whose sensors were not that great, and they couldn’t fix them, so we went with a Taiwanese manufacturer even though they were twice as expensive. Here’s a writeup about those two parts: http://deepfreezefishing.com/blog/2013/09/testing-every-part/

      For PCBA we contacted two local companies I’d toured in the past for a different project. You can just call them up and schedule a meeting with their reps. You’ll get a tour and an hour to show them what you’re working on and talk about volume and expected price. Then they come back with a quote. I also contacted a Chinese company, but things with them broke down pretty quick when they didn’t listen to me tell them not to replace my part with an alternative that clearly wouldn’t have worked. In the end for PCBA we’re doing it ourselves, and managed to get the cost down to about the same as it would be in China, minus the duties, shipping, and QC concerns.

      With the plastics my business partner was already working with them on another product, so that was easy. But they’re close so it was easy to schedule a visit and work with them. Same with the packaging. We’re doing the final assembly ourselves.

      You should be looking for a manufacturer that’s at your level; don’t go for a huge company that does runs in the millions; you can’t afford them because of their overhead, and you won’t have any leverage or attention. But don’t go for the cheapest back-alley option either. You want someone that has the capacity to fulfill your order and give you the attention you need, and can grow to meet your expected demand. When just starting, local is nice because you can go in at a moment’s notice to fix a problem, or get in their face, or even just because shipping is cheaper.

      Finding someone to design your parts is challenging. That’s mostly word of mouth, so find someone close, meet with them, and if they can’t solve your problem, they probably know someone who can. Lather rinse repeat until you get there.

Leave a Reply to Bob BaddeleyCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.