[HeRos] Last week

After a long week-end, we finally present our result to Alexis and Sam. It was good but not enough to be attractive for common usage. HeRos are able to communicate with a smartphone by WiFi, display the camera, shot trough IR. The gameplay is not enough developed and thus it is our first goal for the beginning of the week.

Three points:

  • Adapt android application design
  • Improved gameplay
  • Create HeRos identity to identify the project (logo, color, sound …)

Samuel gave us a very good advise today in order to correct the inconvenient disconnection between the HeRos and the smartphone. We currently used TCP protocol but maybe during re-send packet we lose the camera then the connection is interrupted (not closed) during 1-3s. We will implement UDP protocol to avoid re-send packet since we need the camera only at the present time.


[Kudly] First deadline


Today we had the first deadline : we prensented our project. It almost respected PSSCS but it’s always missing streaming to server.  We used the wifi module in streaming mode and successed to receive data from server and play it on speaker but we have some issues with it to send data from microphone to the server. We will try to fix it before the last presentation.


[Kudly] Connected teddy bear


Today and yesturday, I worked on http post with our wifi module. Now Kudly is able to download or to upload a file from server. It’s also able to read instruction on a xml web page to control LEDs for example.

We spend some time debug streams reading with the wifi module. In fact when data (one or more) are ready a polling command returns 1, when stream is closed (reading is finished) it returns “command failed” and when the buffer is empty but there is other data that is coming it return 0.  So we looped until the polling return a fail or 1 but it blocked half time in this loop because, according to us, we asked too fast the wifi module so it couldn’t close the stream when there is no following data. To fix it, I add a timeout on this loop.


[Kudly] Wifi, SD card & hug sensors


Since my previous post, I implemented a functionality that saves a web page on SD card. This will able Kudly to dowload songs or other files from server. In order to achieve that, I used Wiconnect API on our AMW006 wifi module.

When I had time, I added hug sensors. In fact, Kudly will have flex and force sensors to interact with the child. They are connected on ADCs on stm32 to take pressure values.

Now I’m working on http post !


[StabiloRose] Code revision and rebuild

Today I had a code revision from one of my team members. We also had a brainstorm of how we should implement a better message exchange in the module i’m working with (Wifi) again. So coding day (or night…).

Tip of the day: draw your code to better understand it, is a very good idea.

That’s it for today.

[StabiloRose] Assembly and Wifi improvements

Today I tested and modified some functions that I had wrote yesterday for the Wifi module. Basicaly, as we were not needing to store Wifi information, we were printing it directly to the shell. Now we have a working mailbox that will be usefull for next week PSSCs and future implementations.

Also, we had started to effectively think and experiment the assembly of the sphere. We fixed the 3d printed parts with glue and cut useless parts from the external sphere. We also tried  to fix the leds at the surface of the internal ball witch made us think that the best solution for the fixation of the led strips will be the idea of concentric circles (our led strips are not flexible enough for the orange skin approach) .

That’s it for today.

[StabiloRose] Some more modules are alive

Last Friday and Saturday (03/03 – 04/03) I worked at the Wifi module establishing the serial communication and doing tests. Clement helped me when we where blocked at a reception error. I also took a look yesterday at the buzzer branch, witch is now making the buzzer sound and waiting for little modifications for the next week PSSC´s.

Tip for future rosers: Always initiate to code your functions from a well written “init” functions. Problems like software reset (that was our case) can be easily avoided, saving many hours of work.

As our project don’t have many isolated modules, we delegated (for the moment) each part to someone (ledstrip, imu, wifi etc).

“Human Courses” took half of my last week and last night but now I’ll be trying to work at the improvement of wifi code (implementation of mailbox etc) or helping Yves at the IMU code (since we have loads of PSSCs linked to that one)

That´s it.

And Tip number 2: Always doubt cables and jumpers!

[Kudly] Wifi & integration


Today, I finalized to implement hhtp request and parsing of xml frame to light LEDs. Now, we can regularly send requests to our server and execute actions written in xml in a web site.

For integration with others features, we had some issues. In fact, our wifi module communicates with stm32 in UART (we are using it) but also in SPI. The problem was the wifi module did a factory reset every time that we flashed the board. On the SPI bus, there is also SD Card module. By default, on wifi module SPI MISO pin is assigned by factory reset pin. When we connected the SD Card module this pin was set and wifi module did a factory reset during the stm32 reset time … To fix it, we configured this pin in input for wifi module.


[Kudly] Wifi & leds

Hello !

This weekend, I linked the leds with Chibios Shell. We can now set colors to leds using shell commands.

With Damien, we are going on with the wifi. Now we can make a request to the server to get a XML file that sets the led colors. We are going focus on sending a stream now, which may be trickier than simply make http requests with the AMW006 API.

We are probably going to set up a websocket between our board and the herokuapp server in order to receive the stream on the server.

[Kudly] LEDs & wifi


We received our Kudly boards yesturday ! We tested some features as UART debug, shell or PWM. As Dimitri said, there is one card where UART doesn’t work…

Today, I worked with Marc on LEDs PWM and we spent a lot of time to uderstand that LEDs were soldered in reverse and to deal with a complementary timer. After, we began to send some commands to our wifi module to download updates and connect it to the network.

We hope finalize it for tomorrow !

Bye !