Timing of New Product Rollouts

FIRST General Forum
19

Many of the COTS suppliers are former/current participants of FRC teams and their largest/entire market are the FIRST programs.
Unlike popular companies such as McMaster who sells to a much larger market year round, I cant see how the timing could be fixed unless some fundamental shift in how FRC is delivered, changes.

20

I think I recall hearing in the past that FIRST presents game information to select suppliers around the October / November time frame (though I could be wrong). It’s no secret that game-specific products (or at least products that work better in the upcoming game than alternatives) are released most years during the December product releases, which only gives suppliers 1-2 months to design new products as well as stock what they think they’ll need based off the game info. Personally I prefer having these products and priority stocking based on game knowledge because I think the net effect is a benefit to teams, but 1-2 months makes that kind of work impossible unless the supplier has inside info from FIRST that the other primary suppliers do not.

I’m curious what’s stopping FIRST from telling these suppliers this game information in June, or July. Why not give them more time to plan around their stock (which takes quite some time to prepare and ship) and prepare their new products. I have a hard time believing that the game isn’t ready enough by then to give suppliers a basic outline of what to expect, and if it isn’t then FIRST should consider changing their game design process, because it really isn’t that hard.

21

I fail to see how this would make it impossible to use sensors from those companies. It simply means you’ll have to wait until next build season (or really just that year’s offseason events) to use those sensors. If a product isn’t released until March or April of a year, are you really going to rush to get it onto your competition bot anyway? So it’s really only impacting any products released in January or February, and then it’s only delaying their use by one year.

And I don’t see how it impacts recruiting them as sponsors in the slightest. Heck, Analog Devices currently is a FIRST Choice vendor and Texas Instruments used to provide items in the KoP.

22

Let me do my best to explain it. Most companies don’t sell products to FRC and have marketing departments that are larger than a single Billfred (The standard unit of measure for FRC company marketing departments). When they release a product into the wild, it is accompanied by marketing campaigns and trying to find example customers and use cases. This is particularly important for companies releasing sensors, and doubly so for smaller companies that don’t have the market share of larger companies (like the ones I named).

When they do this in the middle of build season, and it happens, then getting in touch with them, particularly as a team that does interesting stuff and suggesting to them: “Hey we’d love to evaluate this product and see what it can do for us, we’re a competitive high school robotics team, do you think you can spare some samples for us in exchange for a white paper, a statement, or beta/alpha feedback???” can be a big win-win for both a team and a company trying to make a name for themselves.

I used AD and TI as examples, they are far from alone, but I figured you would have heard of them and you have. In the last five years or so I’ve helped our students to seek and obtain sponsorships from companies that make LIDARs, cameras, embedded systems, RADARs, and more - many of which you’ve never heard of. Not to mention software we’ve used which I know some of we’ve picked up mid build season because it hasn’t been available prior to that.

A technology sport that hobbles its competitors from using the latest technology… this isn’t a good idea to me. I’ve explained why. I don’t think you’ve done an adequate job of explaining your own views of why we need to limit this and I don’t believe you’ve thought about all the edge cases for the rule you’ve proposed.

23

One of the edges cases that I can see is a company releasing an update or a fix to a product. That product would be new and unannounced before the season.

If your response is just that “updates to products” are allowed then companies could just release an item called “intake” and update it every year to work with the new game.

We should be allowing more things to be used on robots and reducing the number of rules, not adding to them.

24

Thanks, Mike, for starting this discussion. This is the first time I have heard a significant demand for FIRST suppliers to be releasing many parts in April. We’ve gotten opinions on a few new products by releasing them at FIRST Champs in the recent past (Sonic Shifter*, compliant wheels, new hubs, S3 system), but not many.

My overall opinion is that the market should drive when the products get released. This could be driven two significant ways:

  • Loud demands from end users
  • Mandated rules from FIRST

Either way, AndyMark can adjust as needed. While it won’t be easy for us to make this change, it’s possible because we are now large enough to handle that demand. However, I believe that there would be some significant detriments of FIRST new product releases were mandated to be before a certain date. Before I explain my opinion on these detriments of a mandate, I think it’s important to understand some history.

In our early years as a startup company serving only FRC, AndyMark had 3 distinct business cycles through a year:

  1. Product development: March - September
  2. Manufacturing and sourcing: October - January
  3. Fulfillment and customer service: January - April

Once AndyMark got into FTC and FIRST Choice, the cycle changed:

  1. Product Development: March - October
  2. Manufacturing and sourcing: March - January
  3. Fulfillment and customer service: September - October, December - March

Please know that all of these tasks are worked on throughout the year, but the dates above are the times we are more focused on those tasks.

OK, now I can explain my opinion regarding a mandated rule which would require legal FRC products to be available by a certain date.

  • This would be extremely difficult for inspector enforcement. This will lead to more stressful events and poorer customer service to the teams. New rules which are difficult to enforce are not inspirational for anyone.
  • Startup businesses catering to FRC would have a more difficult task. Within startups, the same people who develop and source the parts are also packing and shipping boxes. There is minimal time to develop products between December and April, as they are supporting team shipments and answering emails and phone calls. Startups would have yet another challenge to succeed.
  • Product costs would increase 5-10%. Pushing back the product development cycle would require manufacturing, sourcing, and inventory to be shifted earlier, 6-8 months before the purchases would be able to pay for these costs. This delay costs money in cash flow, increased inventory, and warehousing.

Another reason not to have this mandate is because there still is a need for game-dependent products to be released along side the new FRC game:

  • Game pieces
  • Game-specific mechanisms - Many teams benefited from the assemblies created by FIRST vendors.
  • There are always surprise parts, added after kickoff, which the teams desperately need. Examples are springs (Steamworks), cRIO connectors and gaskets, 3M reflective tape, high demand Fisher-price motors, slick wheels (and products which dealt with them), unique parts in the kit, and others.

Since we are hearing this new request for products to be released much earlier, I don’t know if it’s a few concerned voices or if it’s a majority. This is something worth investigating. If it’s a majority opinion, then I will push AndyMark to respond to the loud market demand of releasing many products in April (or early summer) in order to meet that demand. Please, let’s not push this to be a mandated policy. Let’s let the free market work it’s magic.

Andy B.

    • When this was released this new shifter many years back, were this close to naming it “Taylor Shift”. I was on board with it, and a few people talked me off the ledge with this about an hour before we released the new name.
2 Likes
25

Mandating a product availability deadline doesn’t make sense. FIRST doesn’t need to save teams from themselves to prevent them from designing stuff that uses unavailable parts. Teams have the ability to make choices to use or not use the new stuff that comes out each year.

I’m sure any FRC supplier would much rather have products available in December instead of late January. They are motivated to sell more stuff this year instead of waiting until next year. Make it happen and take our money.

26

For what it’s worth, my opinion is that this is just some loud voices, not a majority opinion (at this time). Most teams don’t have the schedule or resources to take advantage of product releases that early - Many are season-only teams, some are fall+season, and relatively few meet year-round. Not to mention many teams are buy-as-needed, not stock-up. HOWEVER, if new products were released immediately following champs, I could see it helping teams make the push to the next level - they come off champs excited, see new products that would have been awesome for something on their robot, and make a push to raise some money, get it and modify their robot for off-season events.

As an alternative idea, what about considering an earlier beta/pre-release period for some (probably not all) new products? Make 10-20 of the product, put them up for limited sale or raffle or something, and ask for feedback. Those teams with the resources to take advantage can buy them several months early and put them through their paces. It’s a win-win-win - the team wins, because they get experience with the new product early, the community wins, seeing the durability and reliability of the product through off-season events, and the company wins by getting feedback early enough to make changes before the full release.

Heck, having the pre-release versions at your booth at champs, with a team-based raffle right there would be totally awesome! It would also give you a chance to show teams some of the product development process - product X looked like this at champs, but 6 months later it now looks different - weight reduction here, strength improvements there, changed gear ratios, whatever. Kind of like you did in your release video when talking about the wheels and how they’ve changed over the past decade+.

1 Like
27

I very much appreciate AndyMark’s responsiveness to community feedback, but I must confess that I don’t really believe markets (especially small, captive ones) have all that much magic.

28

If we as a community pushed for a Champs release, or an early fall release, of new products, I think the business model would last for at most one cycle. Startups like WCP, Ozzy boards, etc. would either get on board with the new cycle and risk being sniped by a larger company before build season (witness, among many other examples, ratcheting attachments for gearboxes, and 775 wiring) or they would wait until December anyway and wow the market right before build season. Vex and AndyMark would quickly revert to the December releases to stay relevant.

I’m okay with purchasing for stuff I know we will need every year (such as bearings, raw stock, motors) in the fall, maybe popping for a few wow items once the game is released, but not relying on them as critical parts of the robot design. Last year we bought a Greyt elevator but ultimately chose not to use it, since we had a student design ready to go before the Greyt product arrived.

1 Like
29

Thanks for the reply!

I honestly don’t know how significant the demand is. I replied with what I thought the ideal timing would be from a customer perspective, in a vacuum, and others seem to agree. But I want to add some nuance to my answer, because “just release it all in April” is probably asking too much of suppliers for too little benefit to teams. So let me explain.

As I see it, these are the main reasons why a team may not use/be satisfied with a new product:

  1. Poor (or uncertain) product availability.
  2. Not a good fit for the game/robot (ex. Limelight in 2018 for the most part)
  3. Too expensive.
  4. Unknowns / risk aversion (ex. first CAN motor controllers)
  5. Known product defects (ex. bad batch of hex shaft)
  6. Don’t know how to use it effectively (ex. overloaded blown-up planetary gearboxes)

My assertion is that when all 6 of these reasons are mitigated, the vendor will have a hit product on their hands, and teams will derive value from using the product. Thus it is in everyone’s best interest to resolve each issue!

Obviously release timing isn’t the only lever that vendors have for addressing concerns, but it affects several of these issues:

  • Most obviously, (1) is directly mitigated by having stock available in advance of kickoff. The earlier the product is in stock, the better teams will feel about their ability to obtain it.

  • (2) and (3) are not really affected by product timing. If anything, releasing closer to kickoff probably lets vendors have a stronger idea of what demand will be, and what teams will be willing to pay.

  • (4) is mitigated by getting the product into peoples’ hands and onto robots well in advance of the season. The more “mileage” a product gets in competition settings, the more the masses will have their fears mitigated.

  • Another potential outcome of giving products to teams early is early discovery of instances of (5) and (6). I’d argue that this is in both vendors’ and teams’ best interest. If a product has major problems, the supplier can pull it, improve it, discount it, or at the very least disclose it. If the problems are related to misuse, the supplier can educate about it. Teams will tend to forgive problems that are flagged before kickoff. Less so when they surface late in build or competition season.

Some products don’t really carry a lot of risk. Nobody stays awake at night because a well-known vendor releases a new tooth count 20DP gear or a slightly revised wheel, etc. We all know how they work, and how to incorporate them into our robots. As long as stock is available some time before kickoff, just about everyone will be happy. I don’t think they need to be in my hands in April.

Other products do carry risk. For example, motor controllers are really complicated and have many potential failure modes. Going all-in on a new motor controller is a significant investment and the decision may not be easily reversible in-season in some cases. Many teams that transitioned to CAN in its first year of availability can speak to this. In this case, releasing much earlier can really help vendors address the last 3 bullets.

And to be sure, I think the perceived risk of a given product is tied to the vendor. When AndyMark releases a new shifting gearbox, I am 99.9% sure it will work flawlessly. When a vendor I have never heard of releases one, I am more skeptical at first. Good vendors know this, and go to great lengths to assuage fears - beta testing with teams, demos at World Champs, etc. Smaller vendors also elicit a suspicion about their supply chain and ability to keep things in stock - if the big suppliers struggle with logistics, teams are naturally going to be skeptical about newcomers’ ability to deliver at scale. (Big vendors who consistently have logistics issues arise the same concerns, of course)

TL;DR, the riskier the product and vendor, the more value teams get from an earlier release.

30

If you want your products to be memorable for FRC use they really do have to be announced either in April at or just after champs or in December right before build season. I don’t care a ton personally when the products are announced. I just know that I kinda forgot about both of the FTC product launches until they were brought up during the December launches. The thing I would care the most about is products being available on the day or at worst within a week of announcement. Usually in the past we’ve tried to stock up on parts that we know we’ll use regardless a little bit before season but most game and robot specific parts aren’t ordered until we know that we’ll need them. And generally most of our year specific parts orders are fairly early into build season. We design with what’s in stock and don’t like taking risks on when something critical to our design will show up. While all these new releases are cool and would be nice, if they aren’t available by usually the middle of week 2 at the latest, they probably won’t be used in our designs. It is more important to get a robot working quickly and start testing and tuning it than design with fancy new stuff that solves some design problems we’ve had in the past. I do like the current time frame of the announcements because it gives enough time to look over the items and get ideas of where they’d be useful in designs but I’d prefer if every announced product was available to be shipped out in no more than a week from announcement. (It would also be nice if some stuff would stay in stock but that is probably always going to be a problem because the game changes every year and even if you knew the full game it’s impossible to predict what the primary design will be and prepare for it)

31

If I had my druthers, new FRC products would come out in August/September.

That would allow teams like ours to test products at offseason events for potential use the coming season. It would allow us time to get to know new or new-to-us technologies (CAN, increased sensor or camera abilities, etc.) and gain a level of comfort before build season begins in earnest. It would give us some meaningful project ideas to work on in the offseason. And it would give the vendors time to refine their designs and source their materials/fabricators as compared to an April reveal. As has been mentioned before, this is a game of Top Of Mind Awareness, and as cool as it is to see new products at the vendor booth at CMP, they run the risk of getting forgotten by the time the next school year rolls around.

For reference, our team runs September through April. We do not do any organized activities in the summer. In the fall, we generally meet once a week for 90 minutes.

32

I agree with everything you said above, but I’m not so sure about this. For example, when the EVO first came out it used a dog gear that was prone to failure. AndyMark figured this out relatively quickly, released a new and improved dog gear, and most teams got the replacement in without a ton of hassle. That being said, the announcement was posted on February 21, 2017, meaning teams had already pretty much finished their robots and had to waste either bag-day or event set-up time fixing their transmissions.

The EVOs appeared on the AM website in late December (they weren’t there 12/17 and were there 1/14). That means it took about 2 months for teams to find and for AndyMark to fix the problem. So if the gearboxes had been released in April/May, teams would have had 4x as long to find the problem in the off-season and AndyMark could have implemented the fix even before kickoff.

Obviously this phenomenon is even more true for smaller and less trusted suppliers than well-respected and experienced sponsors like AndyMark. But that doesn’t mean that AndyMark, Vex, and other major FRC suppliers wouldn’t also benefit from earlier product releases.

33

I fully understand this.

And this is fully possible even with the proposed rule change. Do you think this company is going to particularly care that you used the sensor in a summer off-season event or a development machine as opposed to a bot that competed in a tournament that (in this hypothetical you’ve established) they’ve never heard of before? Just because it doesn’t make it onto your competition bot immediately doesn’t mean that you can’t pursue using it at later events or in future seasons. It doesn’t inhibit the ability of bringing that company or that technology into the fold.

Awesome! And this rule change wouldn’t impact that.

You’re right, I haven’t done any explaining in this thread. As I mentioned in my first post, that conversation has already occurred in multiple other threads (Greyt products, WCP MCC 2016, etc). I didn’t feel like re-hashing those debates in full, especially given their only tangential applicability to this thread’s core discussion. You’ve made your argument about this (admittedly) low scope edge case, I just haven’t been swayed by it. In my estimation (and I don’t expect all to agree with it), I don’t think that delaying immediate competition robot applications of new products from outside vendors released in a two month period is a significant enough reason to discard this proposed rule change.

34

Allen summed up my thoughts nicely with this one statement. Rules that limit access to COTS items are bad for the program in my opinion. Know your risk factors and have a plan B if and when something fails.

35

The language can be tweaked to ensure the intent is made clear. More tha anything, this would be a rule put in place to limit the actions of vendors more than teams. It should be abundantly clear which vendors are attempting to circumvent the intent of the rule with products like “intake” that gets redefined each season. The intent of the rule should be more important than trying to find ways to circumvent it.

It would be easier to reduce the rules if people didn’t find ways to break the spirit of rules without breaking the letter.
I agree that we shoudl allow more things to be used on the robot, I just disagree regarding where that dial should be set regarding purpose-built mechanisms designed after kickoff. I’ll leave futhering that discussion to the appropriate threads.

36

That’s not enough. Ok. So what would be enough?

The loophole you’ve left open with the word “announced”? Would that be enough? What about products like the chassis for the kit which has been announced we are getting a new one? We don’t know what it will look like so I’d argue it isn’t fully announced personally - it doesn’t have a SKU right now to my knowledge - AM14U19 maybe - without a SKU, is it a product? Is it announced?

If a company makes an announcement in a forest but no one is around to hear it, is that an announcement?

How about other items included in the kit that aren’t announced yet? Game pieces? Ohh, this rule only applies to things not in the kit, of course you’d say that!

So how do inspectors check for if something was announced prior to kickoff? Do they trust the press release from the company or the kickstarter? Do they leverage a presentation video from a conference (Looking at you RC/WCP!)?

What about products that are announced but then delivered with unannounced features? What about products that are announced with features but then delivered without them?

Where’s the line? Write the rule. I’ll happily pick it apart for you because that’s exactly what teams are going to do with it too.

37

Teams should only be allowed to use what is in the KOP.

38

I definitely get the intent. Unfortunately, the rules apply to teams, and only teams. They are the ones responsible for demonstrating compliance at inspection. And so even if the rules are written to encourage/force vendors into a particular direction, if a vendor fails to do that the punishment falls to the teams with unknowingly illegal items and the inspectors that have to enforce the rules. That’s how it works with the current “Vendor” definition, which is why they aren’t enforced as strictly as written.

So more rules becomes more rules for teams and inspectors, even if they’re intended to limit vendors. I don’t realistically see a way around this. We can move back to a completely prescriptive legality for the whole robot (e.g. list of all legal components, everything not listed is illegal, like we do with motors and control system parts), and that forces vendors to submit all parts to FIRST for legality consideration by a particular date, with supporting documentation on stocking, etc, but that seems like an extreme solution to the problem, to say the least.