smart table tennis racket

[simon_j]

Hi,

I am Simon, the co-founder of janova a young tabletennis startup from Germany. We just launched our Kickstarter campaign for our smart racket. https://www.kickstarter.com/projects/janova/janova-smart-racket/

I have been playing tabletennis for more than 20 years at Post SV Mühlhausen and seen them reach the German Bundesliga. For me, as a player and Fan, that was one of my most memorable moments! I am super excited about table tennis and want to give something to the community. I would like to hear your thoughts on our project! What do you think?

best regards,

Simon

 

[Lazer]

Absolutely totally useless...

Cheers
L-zr

 

[Haraold]

interesting to say the least!

 

[Creek]

I think your general idea will find some crazy people, who want to know more about their game, but from my point of view you forgot about anessential product marketing thing: have your target group in mind when developing a product.
your target group are crazy table tennis people, who focus an details. This means, that most of them also care a lot about their equipment. Having that in mind, those people are are either EJing (switching from equipment to equipment to find the right one, which never ends…) or they looked a lot for the right equipment, found it and do not want to change. This means, that both groups are unlikely to switch to a certain blade just for analysis. E.g. I won’t change to your 7 ply blade or Limba alc blade if I like my Viscaria or like to switch a lot.
in the end I think, that your product will have much more success, if you make a very tiny and low weight unit, that could be added to the end of the handle of any blade. If that’s technically not possible I see very limited users,

 

[Wrighty67]

Fair point raised by Creek but I don't think this would rule out people using a product like this as long as the blade was a good one and potentially at a later stage came in different variants modelled on the popular blades. I also think that it might be that the blade would only be used for certain sessions and to analyse a stroke or to measure improvement after specific training and so would not have to replace the individuals preferred blade. I also think that the "crazy" people referred to might well consist of most of this (and other) forums given the intricate details discussed around equipment and various strokes.

From my perspective, the addition of data to support training is something that almost all sports have integrated and benefitted from - swimming, cycling, running, rugby to name a few - I have used data extensively in my cycling career in the past through Power Tap hubs and SRM units built into wheels or chain sets and it was transformative in improving training efficiency, and so results. I have also used data as a part of my coaching business in order to evaluate and monitor remote client performance - another possible use for this technology instead of or in addition to video that is so often asked for on the forum when or before offering advice (Imagine a video of a player along with stroke data provided from the actual session)

I think the addition of data to TT, given the speed and millimetres precision required, would be a very useful tool to analyse and make changes and I would certainly be keen to use it.

 

[Takkyu_ga_suki]

Can this product give racket head speed at contact and/or throughout the swing?

 

[brokenball]

Simon_j, you are talking to a relatively unsophisticated audience. You need to explain more.

This product can be good, but I wouldn't call it a smart racket. It is an "instrumented" or "sensor" racket. I bet most people here don't even know what you are trying to accomplish. I have some questions.

How is it powered?
If is battery powered, then how long will the battery last?
How fast is the mems chip? How fast can it sample six channels? I know the video said 9 but 3 of them are for magnetic orientation and not necessary.
Where is the data stored?
Where is the data analyzed? This is the hard part.

For the forum.
A product like this is good for analyzing strokes. The accelerometer measure acceleration, what else?
The acceleration can be integrated once to get velocity and against to get position. 3 of the channels measure the orientation of the paddle. With proper math it is possible to recreate a stroke and show it on a computer.

I was waiting for being able to read 6 channels, x,y,z, accelerations yaw, pitch and roll angles every 100 microseconds. Then it is possible to see where in the stroke the ball impacts the paddle.

This isn't useless if it is done right, but it shouldn't be called a smart paddle and there needs to be more than just the paddle. There needs to be software that can analyze the data.

 

[DukeGaGa]

I backed it just because lol

I think I’m a different type of EJ, I buy blades mostly just because it looks interesting, not because I want to use them long term. 😂 But if they just happens to be pretty good, I’ll use them long term. Most recent example… bought a Mina ito blade, and also the rubbers she uses. I tested it for maybe half an hour then a friend took it away because she wanted to try out short pips, and after using it for two weeks she’s thinking about buying it from me lol

 

[Michael Zhuang]

I backed it just because lol

I think I’m a different type of EJ, I buy blades mostly just because it looks interesting, not because I want to use them long term. 😂 But if they just happens to be pretty good, I’ll use them long term. Most recent example… bought a Mina ito blade, and also the rubbers she uses. I tested it for maybe half an hour then a friend took it away because she wanted to try out short pips, and after using it for two weeks she’s thinking about buying it from me lol

Im the type of EJ that is looking for really good but underfollowed blades so that I can recommend them to other people.

Recently I have been setting up paddles for several guys, and I like getting them good deals and good setups.

 

[simon_j]

Can this product give racket head speed at contact and/or throughout the swing?

Yes, indeed. Even though we are still in the process of implementing all these features in our app. We are technically able to measure all that.

 

[simon_j]

Hi brokenball,
thank you for this comment. Indeed we are missing some information on our kickstarter-page. We should add them in our FAQs. But let me try to answer your questions.
It will be powered by a battery. Right now, our prototypes are powered by a 100mAh-battery and we never managed to drain the battery during a whole day of training sessions. In standby almost no energy is used, so the battery can last for several weeks or even months.

How fast is the mems chip? How fast can it sample six channels? I know the video said 9 but 3 of them are for magnetic orientation and not necessary.
Where is the data stored?
Where is the data analyzed? This is the hard part.

I have to admit, not all of your assumptions are correct. But that is indeed to much detail for this Forum.
Inside the racket, the data is preprocessed, buffered, and then it is transferred to the smartphone via bluetooth. The final analysis is done there.

For the forum.
A product like this is good for analyzing strokes. The accelerometer measure acceleration, what else?
The acceleration can be integrated once to get velocity and against to get position. 3 of the channels measure the orientation of the paddle. With proper math it is possible to recreate a stroke and show it on a computer.

You could stop there. This might be interesting at first, but quite boring after some time.

I was waiting for being able to read 6 channels, x,y,z, accelerations yaw, pitch and roll angles every 100 microseconds. Then it is possible to see where in the stroke the ball impacts the paddle.

This isn't useless if it is done right, but it shouldn't be called a smart paddle and there needs to be more than just the paddle. There needs to be software that can analyze the data.

You are right, the smart thing is the software. However, the software is split between racket and smartphone.
Also, it is our hope, that the racket together with the app is (or better will be) much more than something that just shows you your current stroke. We are designing our app to be motivating. For us it is all about motivation. For me it already a benefit, if I can measure my intensity during training drills. Now I can really connect to something, when my coach tells me: "do this drill at 70% intensity". I can try to improve may intensity, while at the same time minimising my error-rate. If I just incorporate these two objective measures (intensity/accuracy) in my training and track them over several session, I can really improve. There is so much, that I can take out these "stupid" 9-axis to track my style of playing, which at the end helps me improve my game.
Maybe our racket is not for everyone, but we hope there are some people that share the same passion for table tennis and technology and are as curious as we are to see, what we can do with this in table tennis.

We will keep pushing!

 

[simon_j]

I think your general idea will find some crazy people, who want to know more about their game, but from my point of view you forgot about anessential product marketing thing: have your target group in mind when developing a product.
your target group are crazy table tennis people, who focus an details. This means, that most of them also care a lot about their equipment. Having that in mind, those people are are either EJing (switching from equipment to equipment to find the right one, which never ends…) or they looked a lot for the right equipment, found it and do not want to change. This means, that both groups are unlikely to switch to a certain blade just for analysis. E.g. I won’t change to your 7 ply blade or Limba alc blade if I like my Viscaria or like to switch a lot.
in the end I think, that your product will have much more success, if you make a very tiny and low weight unit, that could be added to the end of the handle of any blade. If that’s technically not possible I see very limited users,

Hey Creek,
we actually considered "low weight unit add on" option, but at the end it didn't feel right to put something onto your racket. It just felt cleaner to have it hidden away inside of your racket.
Interesting story about the changing equipment. Before we started collaborating with Tibhar we were also playing different equipment. In our team we had Donic and Butterfly players - and we haven't switched for the last 20 years. Also, if it hadn't been for our project, we probably would have never thought of changing anything. At the end we tried different Tibhar blades and we really liked the DJ7 so we also chose it as a basis for our racket. After this experience, I must admit that switching gear is not as difficult as I originally thought.
The second thing is that we as Janova plan to integrate our technology in different table tennis brands. Right now, we collaborate with Tibhar, but are already talking to other brands. We will see how that will go, so maybe no-one has to change at the end

 

[lasta]

I think this could be an alternative to the expensive video analyses stuff.

But, spend some time here and you will see that racket preferences are all over the place. Rather than a "smart racket", I'd like to see this as an "attachable" to any racket.

 

[simonstuetzer]

The idea is to have a smart version a different rackets from different brands. It doesn't feel consistent to put something on the racket and may not deliver the optimal results.

 

[brokenball]

I have to admit, not all of your assumptions are correct. But that is indeed to much detail for this Forum.

What assumptions did I make that are incorrect? I get extremely pissed when someone just tells me I am wrong without justification. I was mainly asking questions.

Inside the racket, the data is preprocessed, buffered, and then it is transferred to the smartphone via bluetooth. The final analysis is done there.

This seems reasonable but you avoided answering my question about the sample rate.
How much data is stored? There are 6 mems channels that are important. No one is going to care about where magnetic north is.

You could stop there. This might be interesting at first, but quite boring after some time.

I am a geek and understand these things. I have used accelerometers in real projects.

You are right, the smart thing is the software. However, the software is split between racket and smartphone.

This makes sense. The smart phone would need to do the complicated math. Display on a regular computer screen or lap top would be a big plus.

Also, it is our hope, that the racket together with the app is (or better will be) much more than something that just shows you your current stroke.

How does it show the current stroke?
I think it is OK to store just one stroke at a time since there is a lot of data to buffer.
There are 6 channels that are important. If each channel samples every millisecond for 0.2 second for the stroke then there will be a need to store 6 x 200 samples. This can be done easily in internal memory of many micro controllers. I would like the mems unit to be able to update every 100 ns so is is possible to get a better idea of the contact ( dwell ) time. So far I have not found a mems unit that will update every 100 ns or 10KHz.

We are designing our app to be motivating.

Now you are losing me. I want info. If an accelerometer is fast enough, it should be able to determine where in the stroke the ball made contact with the blade. This is important. The rate of change of paddle attitude at this point is important.

Maybe our racket is not for everyone,

Probably not, but if implemented correctly, it would be a big help to everyone.

I was thinking of doing this about 3 years ago but the project got killed by an instructor. I was going to mount the recording device on the back side of a paddle since the rubber and the recording device weighed about 40 g each so the weight would not change.
The problem I ran into was that the mems device would only update about every 10 ms which I think is way too slow to detect when the ball impacts the paddle and for how long.

 

[palguay]

Hi Simon,
First congrats on getting to kickstarter, I am also someone who is interested in pushing my Table tennis learning further by using technology.I understand how hard it is to develop something as well as getting it out in front of people.

I have developed my own Table Tennis robot which I have been shipping over the world for the past year, as well as been in contact with other folks like fastpong, osai , stupa as well as chorei king.

I was one of the earliest adopters of chorei king (same as your product), I used it for a few days and the product was new at that stage and I felt it required a lot more polishing as well as not something useful for me at that time. I agree with Creek and I had given the same suggestion to Steve of chorei king that it would be better if there was a device that we can fix to any blade because as a TT player and a bit of an EJ I am very particular about the type of blade that I use and it took me a while before I could settle on one blade. Also the benefit of a bigger market in terms of business if you do not go with your own blade.

I think your product is one part of the AI puzzle, if you are able to integrate it with a video analysis, I meant combine it with the analytics of stupa or osai this will be beneficial for getting a lot more information about the match and you might be onto something if you are able to get the analysis right.

If you are interested in more details , connect with me through the details on the PongFox website and I will be happy to talk through some of my learnings.

Best wishes,
Kiran

 

[OldUser]

Absolutely totally useless...

Cheers
L-zr

You don't explain why it could be useless. Absolutely totally useless answer.

In many sports now sensors are used for analysis, for example in rugby France's coach Fabien Galthié (France won the 6 Nations tournament by completing Grand Slam last saturday) recently said that the most important guy in his staff was Nicolas Buffa, the analysis manager, because everyone was relying on his informations, either the defensive manager Shaun Edwards or the offensive one Laurent Labit, or the most specific managers like the ball conquest or the scrums/rucks ones. Of course, putting light sensors like GPS in rugbymen's shirts does not affect that much their performances, it's probably a different problem in a blade's handle.

To me it could be a nice addition when the pros are in training/camp mode, with the coaches analysing their stats more deeply.