Maelstrom In Your Home Via Opensource

Well, it’s been a while, and updates on information and firmware have all but stopped. The short reason is that basically trying to do things as legitimate as possible has been killing the product slowly — and my wallet which is the primary fund of Titan Lab. I hit this realization a few months ago and advice was to shelve it and try again after a different product — something that doesn’t plug into an outlet! But I said I was going to open source it. I am holding to that. So the code is here:

https://github.com/kwakeham/Maelstrom

Lets start with a status of firmware. Generally people think heartrate would work better. My findings are no (yes, but mainly no). It’s easy to think that heart rate solves the interval problem. For some people it can work, but basically it’s requires constant fiddling to get it to work and for some people it’s a mess (Me). Non-linear ramp, changing high and low points, etc. Can it work, I’m sure it can. Can I make an algorithm that works for everyone — I am doubtful without a lot more resources and data. So this will be open-sourced to the GitHub.

Hardware is another matter. Basically I don’t have a straight answer on some things. The requirement of an airgap. With IEC plugs I can unplug thus creating an airgap, but basically it means a mechanical switch. So that requires a larger (more expensive) IEC socket or a separate switch to meet full CSA requirements. So the design needs a minor update.

So what’s happened to cause this decision now. It’s all business, money, manpower, insurance, demand, etc. I had a couple of people willing to help on Maelstrom but it’s basically evaporated due to pandemic (one refocus on job existing job as others left, one joined Garmin, others had kids to deal with). This left me alone again on technical. It’s just been a financial drain behind the scene and not cheap. So I want to cut my losses as I move on to something else. That something else was going to be sort of related to Maelstrom but that aspect is gone now. It’s a stand alone bike product and I’ve basically finished paying for everything. Tooling, parts, equipment, etc. It’s been tested for months in Canada and the Netherlands. I have a firmware expert helping me and had more interest in helping with the project. So more help, more money, bigger market, lower certification costs. And AMAZING BUTTON FEEL.

So some of the questions I’ve gotten leads into the numbers.

Why don’t you *just* crowdfund it?
The glory days of an inexperienced nobody with zero engineering background and a few renders and a $500 dollar video are over. Talking with the local Calgary startup community had little expertise, and reaching further into my network those who had witnessed it said that up front their campaigns needed 30k USD. Can be cheaper but up front 30k. Why? Most of them didn’t know how to market, so they hired consultants (so would I). Being a consultant (on engineering, strain gages, sports tech, etc) I know that’s expensive. So internally if they hired the expertise it’d cost maybe 50/hr but that comes with challenges. So via consulting it’s 150-200/hr and you better believe every hour is billed! So you also need a video to go with that marketing expert. I make videos but I’m not good enough for a Kickstarter video and marketing effort. The startup consensus is that it ranged from 10k – 20k USD for involvement of a proper commercial venture on the cheap and it can really backfire / look poor / etc. So spend more here. In a lot of cases they want actors, locations, and script writing. So you now have a bill that is easily 30k for a niche product that basically sells to Zwift and Trainer Road users. Going through the math for certification, testing, etc, AND nobody gets a salary, I estimated I need 100k just from the Kickstarter. So 130k when I factor in that up front cost of marketing (not including the thousands I’ve spent to date!), no salary and about 1300 pre-orders at 100 USD each needed becomes hard to swallow in light of the technical reasons.

While there are trainers and accessories that plug into the wall in the cycling world that somehow are not UL/CSA certified, ALL advice is don’t even try. So in consulting with professional engineers who deal with certifications, in the end after thousands of dollars for advice and documents for standards I was met with is this consensus: “Might work, might not. Never seen anything modern certified to this”. Ya, that’s paraphrased, but you get the idea. So even with using a pre-certified FCC module, because it plugs into the wall the whole thing needs full recertification and not just the light battery operated low power wireless. Full blown complex stuff. This is not 10k as I’ve normally experienced but more like 20k… for one country. 30k for two if I want expertise if it fails and help resolving (which I’d need). And that is not including the more expensive/difficult UL/CSA. Those are the non-guarantee pay up front of 30k for CAN/US and basically told to find a company in the EU and one in Australia to sort out their certification. EU advice ranged from CE is pretty loose to your gonna have them all confiscated on shipment. No straight answer except the cost. To Add back the EU, UK and AUS market has grown to estimates of 100k when employing consultants to push things through.

Sum up
FCC 30k
UL/CSA 30k
+UK/EU/AUS 90k

So what? If it’s profitable, who cares. I’ve already have had investment offers (not big enough though). So how many do I need to sell to break even?

Cost of goods calculation.
Because I decided to use things like a safe fully isolated AC/DC power converter (rare in consumer electronics sub 1000 dollars) and opto-isolated triac control (common) and the use of IEC connection + fuse + expensive rare pigtail cables meant my cost of goods was at 60 dollars per unit (CAD) for 1000 units. ECK. A 3x to 5x Cost of goods to MSRP is normally used. I already set the approximate price (incorrectly as a marketing whiz tells me) at 70 dollar USD (100 CAD) so that needs to change. In order to drive views to the site to get pre-pre-order sign ups it cost me $30 (CAD). So 10 dollar CAD margin isn’t going to work for staff, support, building rentals, insurance, etc. So I need to lower the cost somewhere and/or raise price. But I need it to be much less than the one device in the market that does something similar.

And this is where the numbers can’t be pushed to work in anyway without some “dangers”. Startups are risks. BUT calculated risks. So I start with what I can do. I have to change the price. One way is to decrease safety — NOPE. Okay, the other way? Remove the IEC Jacks. WAIT! If you do that you cut out 60% of the market according to your hundreds of pre-order signups. I cut the market in half, the certification in half, and therefore sale units in half. And how much do I save. $10. Maybe $15. So 45 COGS -> 149 CAD price or about 129 USD. This almost works for math. ALMOST.

Up front = Kickstarter, FCC, UL/CSA = 90K. So Kickstarter now needs to generate about 200k min. That’s 1550 units. So after that’s all said an done, what do I have potential for (by that time) 2 years work. I have 34k USD there for everything I didn’t include (Kickstarter fees, rent, test equipment, warranty, support personnel, logistics help, etc, etc).

If someone was willing to put 500-700k into this to scale up then long term the math shows it could be profitable and I have market data. However, it’s been over a year, and I couldn’t make it work, therefore open sourcing it.

I’m kind of heartbroken about it. I honestly did this one first because I thought it was the low hanging fruit on my list of startup ideas. It was suppose to be the easy thing. I overlooked certification and how much things cost and how uncertain / different enough that there weren’t really experts. When you’re not the expert things get expensive. The other side is the loss of people interested in helping. While I’m glad the next thing has newer more invested people from financial/marketing/engineering, I’m saddened that it didn’t catch with other engineering type people to want to be involved.

When I first tried the first prototype it was a 4/10 on my “this is cool scale”. Fixing the triac timing math brought it to a 5/10 or 6/10. Take it away while indoor riding and it’s an 8/10 device. I still use mine when I ride inside (not summer though).

So the what’s next. It’s an accessory but unlike Maelstrom starting around the 4/10 “cool scale” for me, this started at 7/10. After refinement and testing and feature adds it’s closer to a 9/10.