Tennis score board


(Jurjen de Vries) #1

The tennis score board is intended to automate the score publication system for small scale tennis competitions.

challenge: During an successful hackathon at the Internet of Things meetup in Utrecht a couple of teams have designed a prototype for this system. Now there remains an challenge to finish this scoreboard before the deadline of 14 October 2017. If you like to help you can find the Github page at https://github.com/gruifin/Tennis_scoreboard and Slack channel at https://thethingsnetwork.slack.com/messages/C72VA0ZPA . Of course you can post here as well.


#2

If you want people to help you and your team, it would be nice to have more detailed information.
Your github is empty and here I don't see circuit drawing(s), componentlist or explained what kind of help you need.


(Jurjen de Vries) #3

Good questions BoRRoZ. It is because of we just started, with a couple of hours during the Meetup with an introduction from Heerd Jan, who is helping the tennis tournament for children in Woerden. They like to see a modification of their score board so they can see the score online. The current score board looks like this:

The idea is to add tilt sensors to each yellow and red flip boards to make this happen. Then we can connect the tilt sensors to an The Things Uno, and from thereon move the data to for example Node Red > Website.

After the introduction of tennis Ludo started to present about TTN and we created small groups to find out about connecting the tilt sensors and measuring it on the Arduino. Another group was investigating how to install Node Red and how it is working. After a couple of hours we decided to move the project to an online environment. So now that is going on. I expect that some of the attendees will put their code on Github soon, so that this will be visible for you and others.

We can use help with Arduino code optimization, suggestions how to make the measures of the tilt sensors visible on a website, if possible without hosting costs for things as Node Red, maybe it can be done by IFTTT as well or so. and help with coding the CSS/HTML to show the scoreboard and the position of the flips.

Hope this make more clear.


#4

thank you for the explanation .. nice idea.


#5

I see 3 parts

  • mechanical / contacts
  • electronics / wiring
  • software / LoRaWANpart / reading contacts

for reading contacts (up to 16) I would suggest I2C mcp23017 port expander

every position of the switches fits and correlates to two bytes

11001101 000001 xx
these two bytes can be transmitted over LoRaWAN and interpreted in your application

for the switches you can think of reed switches so thereā€™s no mechanical connection
just mount a little magnet on each flipper.

for the LoRaWAN part and as a ā€˜main processorā€™ I would use the TTN UNO ā€¦ you can use that board for reading the contacts over I2C (and a good API)

attention points :

  • is there a gateway in range of the scoreboard / on tennisclub ?
  • should the system operate on battery


mcp23017 in a joystick/debounce experiment :sunglasses:


(Boris) #6

Good stuff BoRRoz

on your attention points:
LoRaWan coverage should be ok. Gateway is near. To be tested soon though
The system should operate stand alone. So battery operated indeed


#7

I looked at your setup code for the tiltsensors.

You use continue polling for all the contacts (loop) , if something changes then the new status is send, that is for a battery operated unit not very power efficient.

  • Better would beā€¦ if nothing changes continue sleepingā€¦ only wake up when neccessary (imho)

If you want to stick with tiltsensors, then you can also make use of this MCP23017
That chip can generate an IRQ when something on one of its ports changes.
That IRQ can then be used to wake up the processor, read the status of the switches and transmit the LoRaWAN packetā€¦ then go back to sleep


(Boris) #8

IRQ driven would be more power efficient indeed. Looks like a simple approach. I guess the sensors have to be connected to both the MCP23017 and the Uno.

I have no view on the actual power consumption of 14 tilt sensors and the polling approach.


#9

the switches are all connected to the chipā€¦ which is connected over I2C to the UNO + 1 IRQ line + VCC/GND

  • what is missing in your code setup is the debouncing for the tiltswitches
    see examples