Tiny Accelerometer to be placed inside ball

Thread Starter

Chillsoffear

Joined Jul 21, 2022
7
Hello smart people.

Happy that I found this forum. I am a Ui/Ux Designer interested in creating a product for a hackathon. In the end, I will have a team that will help me.

Until then I want to ask for help in defining my concept/product.

I want to create a sensor based on an accelerometer that will be placed inside a ball. My problem is that I don't know what kind of parts I need except the accelerometer. The sensor should record data and have an 8h battery life to ensure that it is recording data for an entire training process. Also because the sensor is inside the ball I want to communicate somehow with it to download data. The sensor should be tiny, please see the video below:


Thank you.
 

Ya’akov

Joined Jan 27, 2019
9,079
Welcome to AAC.

I am sure we can help you but we will need more information before that is possible.

While you have said something about the a problem you are trying to solve, it is not the one you need our help with. We need to know about that, in particular, what problem will the sensor in the ball solve.

That is, why do you want to doing this, what will the sensor do that satisfies some requirement? To that end could you explain what you expect the sensor do inside the ball? What is the problem a sensor in a ball is going to solve. That's a baseline so we aren't operating in the pulling teeth mode of information gathering. A user story would be nice.

Then:

What is your budget?
What is your timeline?
What skillset will your team have?
Are there constraints that we wouldn't know about aside from these? (e.g.: rules)

The basic idea of accelerometer in a (basket?)ball is not a trivial undertaking and may be difficult to pull of in a limited time/budget situation but we can do our best to help.
 

Alec_t

Joined Sep 17, 2013
14,280
What sport? From figures I've seen online a kicked soccer ball undergoes an acceleration of around 300g (I can't verify that). So any electronics inside would have to be able to withstand that acceleration. The power supply is likely to be the heaviest item and the gizmo would need to be quite rugged to restrain it.
 

Thread Starter

Chillsoffear

Joined Jul 21, 2022
7
Hi. The sport is basketball. I am trying to build a platform for trainers/coaches to gather data from 3 different sources: a sensor on the player, a sensor inside the ball, and a drone video for a better view of the player's position on the field. I think that I can make good relational data between the player and the ball during a training session. The data can be saved on the device itself I don't need Bluetooth or wireless communication during the game, because I want to make the data analysis after the game/training is ended. So I need to have a sensor inside the ball with a 4-5 hours life battery that can be wirelessly recharged maybe? and some storage for data. The budget is max. $400 to build a prototype. The timeline is like 1-2 months. Thank you.
 

MrSalts

Joined Apr 2, 2020
2,767
Why not analyze the video with a stereo camera setup. You should be able to estimate the x, y and z acceleration based on video analysis captured by the drone.
What do you plan to do with the accelerometer data? Will you ask player to "improve by reducing x-vector and increasing y-vector trajectories when shooting from 1.34 m from this point on the court"?
 

Thread Starter

Chillsoffear

Joined Jul 21, 2022
7
I don't want to think about how I will place the sensor inside the ball for now. I found this idea from Fifa. For now I want to focus on the sensor creation. Thank you.
 

Attachments

Thread Starter

Chillsoffear

Joined Jul 21, 2022
7
@MrSalts Yes this can be one point: "Will you ask the player to "improve by reducing x-vector and increasing y-vector trajectories when shooting from 1.34 m from this point on the court"?"
 

Ya’akov

Joined Jan 27, 2019
9,079
Hi. The sport is basketball. I am trying to build a platform for trainers/coaches to gather data from 3 different sources: a sensor on the player, a sensor inside the ball, and a drone video for a better view of the player's position on the field. I think that I can make good relational data between the player and the ball during a training session. The data can be saved on the device itself I don't need Bluetooth or wireless communication during the game, because I want to make the data analysis after the game/training is ended. So I need to have a sensor inside the ball with a 4-5 hours life battery that can be wirelessly recharged maybe? and some storage for data. The budget is max. $400 to build a prototype. The timeline is like 1-2 months. Thank you.
TLDR; I recommend dropping the ball (no pun intended) and focusing on deriving as much data as possible from the other instrumentation.

The actual components for make a lightweight, low power datalogger with accelerometer aren't a problem at all. You will be able to find several options. (Tangentially, it might be worth putting a pressure sensor on the board as well.)

What strikes me as the main difficulties in this project are:

Though you want to put it off, mounting the device. It will have to be perfectly centered to avoid making the ball asymmetrical, this seems no easy feet.

If you mount it at the center, wireless charging will be very difficult if not impossible. Wilson chose to go with ultra low power BLE to avoid this problem.

You've set up a situation where success has a low probablity. I don't intend to be discouraging but given your resources what you are describing seems overly ambitious. Using computer vision and body worn sensors seems possible.

But a $400 rechargeable datalogger somehow embedded in a basket ball in a way that doesn't interfere with the normal behaviour of the ball (even this part seems to exceed the available resources) seems far too ambitious in my experiece. Maybe you have more time in the 1-2 months that I am estimating but being very generous on a napkin, there's you and your "team", let's say three people at a total guess.

So four people who somehow manage to spend a very unlikely 20 hours a week, that's 80 person-hours a week, for let's stretch it to 10 weeks and we get 800 person-hours. Of course not everyone can do everything so the person-hours aren't fungible. We need to derate it based on that and the fact that interactions will be necessary also reducing the number.

A generous derating of 50% gives us 400 person-hours to complete a project that includes hardware design, hardware development, prototype testing and debugging, software design, software development, testing and debugging, and then there is still mechanical design and development of the hardest part: getting it into a basketball.

All, by the way, for a maximum of $400 in parts. Most of the parts themselves will be cheap, the number of times the BoM (Bill of Materials is changed during the development and testing will decide how many of the cheap parts will need buying. Then there are the basket balls. They have to be decent ones, and I am guessing you need at least 3, unless they are not otherwise accounted for, they have to come out of the budget.

A $50 each, that's $150 out of your $400 right off the top. Your hardware budget from the electronics perspective is now $250. I think you could get the parts for that, if revision don't require replacing major components.

None of this is intended to discourage you from your project per se, but it is intended to give oyu an idea of what, as a consultant, I would advise my client to scale back on it.

Sensors on the goals: backboards and baskets, the players, something else?? along with your video using computer vision won't tell you the acceleration of the ball but it might be able to derive and more.

You may well have a better idea of what you and your team can do with the time-fiscal budget you have described. There may be errors in my SWAG numbers or things I don't know that make my analysis moot. But from having seen many, many projects fail due to overreach my sincere advice is to can the ball sensor and double down on the computationally derived stuff from the video and sensors placed where they aren't such a vexed project to complete.

I wish you the best of luck. Or course I hope you take my advice but if you don't it would be really cool to be proven wrong.
 

MrChips

Joined Oct 2, 2009
30,720
That is a very interesting project with physical challenges. As Ya'akov said, I think the biggest challenge will be getting the electronics located in the middle of the ball. It would appear to me that you will have to work with a ball manufacturer for this.

If I were doing this I would definitely go with Bluetooth and record the data live. I would use a tiny lithium-ion battery or a super capacitor and have it charged by a piezo-electric transducer every time the ball is bounced (which of course is a lot of charging time and energy). The total electronics can be made very small as all the ICs for this (about four in total) are tiny SMD components.
 

BobTPH

Joined Jun 5, 2013
8,816
I am not sure about this, but I think an accelerometer is not enough to track the ball. Systems that do this typically use a gyro as well so that rotation can be separated from linear acceleration.
 

MrChips

Joined Oct 2, 2009
30,720
That in itself would be a challenge, considering how much the wall deflects when a basketball is bounced firmly on a floor.
That is going to be a good source of energy to charge the supercap.
Also, we would use flex circuits to distribute the weight.
 

Alec_t

Joined Sep 17, 2013
14,280
If you wrap the sensor in soft plastic foam it would be protected; but wouldn't any protection mean that the sensor won't then be subjected to the same acceleration as you are trying to measure?
 
Top