Who and/or what is FrSky and her branding and marketing schemes.

[Konrad]

They bought Jag. to enhance their perceived cachet within the automarket. As I recall this has been a financial blunder. No amount of branding can make a Ford Taurus look like and XKE to any car buyer of financial analyst!

Perceived value is another thing I don't really comprehend. I like to deal with objective variables, like verifiable data point's, performance spec's.

Now to gain street cred takes a lot of hard work. Like when Ford beat Ferrari. Now that's how you make a name for yourself. Make a product that out performs the competition.

 

[MikeB]

So I am confused. I thought that FrSky wanted to keep their new lineup proprietary (both hardware and firmware) so that they could defeat the cloners (i.e. Jumper, RadioMaster, RadioKing, etc). The only way to do this is to NOT allow the innards to be exposed.

But your erskyTx is an open source (GPL?) project and if released for the X20 would expose, (through the source code), all the inner workings and layout of the X20's hardware. Yes?

They have requested that I keep the hardware drivers private, and I can see no reason to publish them.

Mike

 

[Landru]

So I am confused. I thought that FrSky wanted to keep their new lineup proprietary (both hardware and firmware) so that they could defeat the cloners (i.e. Jumper, RadioMaster, RadioKing, etc). The only way to do this is to NOT allow the innards to be exposed.

But your erskyTx is an open source (GPL?) project and if released for the X20 would expose, (through the source code), all the inner workings and layout of the X20's hardware. Yes?

A similar thought had occurred to me.

As I understand it, the X10 became easy pickings for cloners for a confluence of reasons:

• It already supported a popular Open Source OS (OpenTx).
• The availability of the Open Source MPM allowed integration of proprietary RF IP with impunity (and off-loaded RF support issues to a third party).
• It probably didn't hurt that FrSky employees reportedly quit the company taking inside information with them.

The same conditions don't seem to exist for the X20.

In any case, FrSky supports Mike's work, which needless to say, comes with the highest regard.

 

[Landru]

With my work in aerospace, where we dealt with the public's trust, we engineers were both criminally and civilly liable for the documents we signed. You can damn well be sure we did the best we could with the information available. It bothers me no end that corporate managers, MBA's and the like (bean counters) often don't have this work ethic embedded into their culture. If FrSky isn't comfortable enough to put their name on a product, that product shouldn't be sold until the managers have the confidence to put the FrSky name on the product! From here it just looks like corporate greed. FrSky would do well to learn from Boeing what corporate greed will do to the value and reputation of a firm!

You'll get no argument from me about the need for rigorous testing of RC devices.

Whether the current level of testing in the industry is sufficient is an open question in my mind. The recent loss of a model to a bug later identified in firmware likely colours my view. As it happens, that incident didn't occur with FrSky; rather, it was another prominent brand.

The loss gave me insight into the stress testing that the manufacturer in question adopted as a result of the bug. It was significant. Whether other manufacturers have a similar program, I have no idea.

As you pointed out, even in a tightly regulated industry such as aerospace, serious flaws can be released to the wild.

 

[Konrad]

Corporate Issues aside, it is incumbent on us to do proper testing. A classic that is often over looked is the range test. I also burn in all my servos for 5 minutes. I'm still surprised at just how many servos fail inside this 5 minute window. And this with all the development in QA and QC programs and protocols that we now have in industry. Somehow our toy manufactures aren't up to date with these concepts.

Now taking a deep dive into FW/SW and RF issues is beyond the scope, tooling and knowledge of most. I'm truly thankful for firms like Aloft and experts like Mike B that have grabbed these issue by the horns and taken the problems to ground.

 

[Landru]

I'm with you on the need for end-user testing of all new gear.

One element I've added to my initial range test is to ensure that the control link can be recovered after failsafe. (A failure to do so was the bug that caused the lost model.)

And yes, kudos to Aloft and Mike B for their work. It creates an extra layer of confidence.

 

[Wayne]

On that subject, we actually tested and improved how quick Universal ACCST recovers from a failsafe event.

Aloft is also a Hitec dealer. While we don't sell their transmitters, we do sell a number of other items from them, mostly chargers and servos. Talking with the crew there, they had developed a new transmitter a year or two ago, and it was looking good, but when they looked at the numbers to introduce the radio using their traditional methods they decided it was too large of a risk, they estimated it would cost the company one million USD or more. I assume this expense covers deep testing, documentation, spare parts, production and marketing to name a few.

I can tell you that Frsky does not do things using this method, and this allows them to roll out new products for a fraction of the cost.

Konrad I think this is a critical change to how manufacturing works these days for consumer items.

The traditional method probably looked something like this:
Get a rough working unit of new product.
Estimate expenses and potential return on investment and see if management agrees.
Seek feedback from main distributors and other key customers.
Seek funding internally or externally.
Refine product
design packaging, instructions and documentation.
Start Mass Production.
Train Staff and distributors
Stock spare parts
Setup service centers
Marketing
Enjoy many years of sales and service.
I'm sure I'm missing some steps.

Some of this is completely different these days, and some of the major expenses from above are either greatly reduced to just simply gone.

The FrSky method seems to have evolved to something like this for the X20:
Listen to customer base directly from online forums for new product ideas, and come up with with own ideas as needed.
Present rough idea and check for feedback of physical design (CAD rendering)
Develop working samples and share photos for feedback with distributors.
Inhouse testing and refinement. (I have no clue what FrSky does at this step, but we constantly urge them to spend more time.)
Release prototypes to trusted outsiders and service dealers for feedback and testing. (They are now much better on this step and invested a lot of time on this step, the results are a vast improvement of the product at launch.)
Small scale marketing via websites, forums and dealers.
Enter manufacturing of limited scale and sell to public via limited sales channels.
Monitor for any issues and fix as needed.
Full scale production and distribution.
Dealers can stock parts if they like once spares are available, this may be months if the product is in high demand.
And for the first time ever!! Produce a professional manual.

This is just a quick outline for both methods, sure I'm missing a ton of steps, but I can tell you that the Tandem radio was a huge undertaking for FrSky, and while they still have a huge investment in this radio, it would have cost a lot more with traditional methods.

One of the big changes I think is really interesting for some companies is the "investment" step has been flipped around with the use of web sites that the customer's can donate money to a company for them to develop a new product!! I've seen some companies bring in millions using that method. Pretty crazy.

 

[Konrad]

Product development is one thing. And yes this takes a commitment from management to actually commit money to a risk. (Everything new has risk).

I'm thinking of QA and QC programs for products already in their pipe line. Such as statistical sampling, maybe use the TMP (Toyota Manufacturing Process), ISO 9000 quality programs, etc. With the resistor/capacitor swap out problem, the undocumented software interaction, and the high number of Software revision to firmware release all indicate there there is a core problem with quality. These are not easy to address, so they will take a commitment form upper management to fix these and instill a culture that will catch these issues long before they reach the customer. Software may be a simple maintenance item for those of us use to software development. But it is a MAJOR stumbling block to most. Even I was about to throw out FrSky after the false positive report of successful upgrades with the ACCST v2.1.x updates. With these FrSky quality issues, the simple restrictive and over priced Spektrum radios will live on for a long time. From here it looks like there is a lot of low hanging fruit to be pick even without the risk of introducing new products.

As a wiseman said; Quality pays it does not cost!

 

[Wayne]

I totally agree. I just wanted to point out that business models have been changing over the last couple of decades. The old distribution methods are all but gone. The manufacturer/retailer models have shifted dramatically. Many manufacturers/distributors that did not see the changes were taken out of business by staggering debt loads from inventory loans they gave to their retailers as that was the old way to do business. This is the sort of thing that killed most of the major Hobby Distributors.

As for the FrSky weaknesses, I have tried and tried to make changes there.. I simply do not have the sway there that I used to have. When they were a smaller company they listened a bit better. This is typical of any business as it grows, harder to make changes. The other issue is that the owner of FrSky is very loyal. He is loyal to his customers, to his employees and to his distributors to the point of fault sometimes. I think this has caused some of the FrSky issues..

 

[Landru]

It seems to me that there's another factor now that can't be ignored: The role of Open Source hardware and software.

In some cases, Open Source provides a valuable resource that a company would struggle to duplicate privately. Take OpenTx, for example. By all accounts, FrSky is doing a first-rate job in creating a comparable proprietary OS (EthOS), but the cost doesn't seem trivial.

Another example is the MPM project. Like OTX, it's backed by very talented developers and the confidence inspired by a large pool of users. And like OTX, the MPM has also been a market changer -- in some unpredictable ways, I'd say.

 

[Wayne]

The question is can a company rely on Open Source hardware designs for a profit?

Think about it. Lets say ACME incorporated looks at an Open Source project and says, hey, we like this, let's make those, we can help it along, and make a few bucks too. So Acme makes them and gets some sales rolling along and helps make a market for the product. They grow in popularity, and Beta Incorporated notices this and decides to jump into the market. But Beta isn't getting many buyers so they drop the price a little and now they get some sales, but sales start to slow for Acme. Along comes Giganto Heartless and they decide they will flood the market with this same product, but they have a few tricks up their sleeve. They don't need to use those higher tolerance parts, these cheaper ones will do the trick, and we can just dump these babies on the market for way less. Oh boy, customers are happy, they can get the same thing for less money. Sure it craps out 9 months later, but maybe the Acme version would have too. Acme and Beta sales dry up and they are out of the market, or if they are lucky they will be known as the supplier of the good but pricey versions of the product. In the meantime the majority of people buy the cheap one, and then complain when it has issues, hurting the market for this product.

Maybe Acme sees an option to make improvements and releases a new version. But this upsets some of the Open Source team members and they tell people to stick with the original.

Or the Open Source team now sees a lot of success with their design supporting these 3 companies, and they start to wonder, hey, where is our profit? Or they just start to fight with one another over the future of the project. Before you know it you have 4 different forks of the original project and 4 different names, and they all have a little different focus. What do the manufacturers do now? Who do they align with?

Ultimately I think it is a race to lower and lower profits for a company as more and more companies will jump into the market if sales increase. This ends up being a race to cheaper and cheaper versions.

One of the best Open Source projects I ever saw for this hobby was for the CC3D flight controller back in the early days of multirotors. They had open source hardware, software, documentation and a great and helpful forum and web presence. I was asked to get hardware into the US market as it was very rare at the time. I was blown away with how good this team was, they were hitting on all cylinders. At the time people were paying over $100 for the CC3D hardware. Quickly it gained a following and demand for the controller and many manufacturers jumped into the arena. As it really gained popularity the Open Source team imploded, and the hardware got so cheap that I think all profits for manufacturers were gone, and with so many manufacturers in the game, they all saw big sales and produced far too many boards, and this caused even more issues. I think there are still loads of these outdated boards available. The CC3D was a great board, it could have been further developed and kept up to date, but with the team gone, it simply went away replaced by other designs.

 

[Konrad]

MPM is a problem, well here in the USA. As the airwaves are held in the public interest. Those that use the airways (transmit) are subject to the FCC rules. For the 2.4 gHz band this means that ALL transmissions device are subject to 'Type Certification', through proper test and certification procedures. This means that some entity must pay and submit these certifying test to the FCC.

My understanding is that NO MPM has passed these type certification. Yes, one of the on-board protocols has been tested but only at 1/10 the possible power rating. If true this means that no RF output is allowed (certified) in the USA at any operational power level.

Some say so what! Well, the sale of the MPM puts those that have followed the rules and gotten the proper FCC type certifications for their 2.4 gHz transmitters at a disadvantage. As the competition Radiomaster Jumper and TBS have not had to incur the expense of proper and legal certification.

All you have to do is look back at Boeing and her attempts to circumvent the certification process with the FAA. Should the FCC get wind of this it will be uncomfortable for the small firms like RadioMaster, Jumper, TBS should the FCC turn over their finding to the DOJ. One can see this with Boeing, as Boeing lost 50% of her value as a direct result of their shenanigans with the FAA certification process.

So while the MPM looks to be a great idea for a product it has as of this date not been done properly when looking as the protection of the public's interest in the public airways.

Then there is the problem of what happens to us hobbyists should we damage property or injure bystanders using these illegal modules.

The MPM is a good idea, but is just a poor product (legally).

All the best,
Konrad

 

[Landru]

The question is can a company rely on Open Source hardware designs for a profit?

Recent history suggests that's a very good question to ask.

The MPM seems to be especially problematic in terms of market consequences. Personally, I think the premise is questionable and that there were some significant unintended consequences, even for some ardent proponents. Then there are issues of certification as Konrad mentioned, and also of production quality. The topic can raise blood pressures.

 

[Wayne]

Just for clarification, I don't think TBS has ever made an MPM, and pretty sure they have proper FCC certs for their gear.

 

[Landru]

Their Website lists this unit, which contains four different chipsets:

Team BlackSheep Online Store - TBS MPM

4-in-1 multi protocol module for TBS Tango 2 and other lite adapter radios

www.team-blacksheep.com

Designed in collaboration with RM, they say.

I bought a MPM to play around with. It was a Jumper unit in my case. The best use I found for it was to experiment with tuning. Now the universal ACCST firmware goes one better with automatic tuning.

 

[Wayne]

Opps, I stand corrected. Think TBS is US based, or were, this opens them up to liability from the MPM.

Don't think the MPM are legal to sell in the EU, so many people buy them from China. The EU actually are pretty quick to fine companies that are breaking the laws, and they have a lot of laws! Been interesting to see all of the things Horus RC has gotten into trouble with in the EU due to the many odd little things companies have to comply with there.

 

[Landru]

After the hefty fine handed out by the FCC to GetFPV, you'd think US companies would be cautious. Something in the order of $2 million, wasn't it?

 

[Konrad]

I think that was Hobby King for the orange RF decks modules and their orange TX. As I recall HK is flaunting the fine. So this has been handed off to the DOJ.
Would anybody in their right mind, buy a Radiomaster of Jumper if it didn't come with a MPM?

I think you are correct.

FCC Penalizes Marketer of Ham-Band Drone Audio-Visual Transmitters

<span style="color: #000000; font-size: 14pt;">The FCC has imposed a $180,000 civil penalty on a Sarasota, Florida, company that had been marketing noncompliant audio-visual transmitte

qrznow.com

 

[Landru]

Would anybody in their right mind, buy a Radiomaster of Jumper if it didn't come with a MPM?

If it didn't come with an MPM, it wouldn't be a radio.

So I was out by an order of magnitude about GetFPV. Where did I come up with $2 million, I wonder? Maybe they negotiated it down. I recall reading of that happening in at least one case involving RC. Still, $180k is nothing to sneeze at.

 

[Konrad]

Marketing crap can be expensive!

Hobby King was fined the 2 million. We will see what the DOJ collects!