Dear @vjreddi , thank you so much take your valuable time to express your intention. I hope that the cheap ESP32 module (with supportive Arduino plugging) would be able to make compatible with TF Micro platform if every expertise given contribution as team work to share the knowledge. we’ll be able to do anything if we are get together. ultimately this world make better place to every one.
@vjreddi Yes, I guess community feedback also helps. It should enable some kind of cross-pollination between several disciplines.
At this point it seems that after the first milestone the team can post a video summary/demo along with some documentation. The community members can just comment/review that work. Maybe the number of upvotes will indicate how much “expert” is that review.
I find this a wonderful idea! After all, creating a good data set is most of the volume of a ML project right? So, the “design” of the data set should be a part of the competition somehow. Of course if a new data set is created, will not be a “perfect project”, at least in a short term perspective. But for example a evaluation factor could be the scale up potential as data is gathered…
In the whole project side, some factors considered could be the potential impact ( under UN SDGs for example), the potential for business, etc…
Sorry, The United Nations Sustainable Development Goals (THE 17 GOALS | Sustainable Development). Since the idea could be evaluate on the “impact”, this could be used as a “framework” or guide to target the projects. For example, I’ve been working for a few years in my company as a water engineer in containerized drinking water treatment systems (think as compact waterworks system) for deploying in rural areas, or small population cluster where no clean water is available. Im working for SDG 6 (Clean water and sanitation). Our systems are already remote operated and can output several information (eg. equipment status, water flow, pressure, water quality parameters, etc.). Beyond the obvious ML implementations that we could deploy in the System it self, I was thinking apply an edge device in the water collection point for example (a river source, a borehole, etc.) with some water quality inputs (such as conductivity, turbidity) that could “predict” the amount of chemicals and energy that were needed to “clean” the water. This would then feed the treatment system and allow for better operation planning and consequently dropping its costs, yielding a more affordable water treatment cost. Perhaps for the course project it self would be a very complex task… This example I think would be more for a long term implementation.
1 Like
Hi All,
Just wanted to follow-up on this thread. I didn’t drop it. Much appreciate all the time and input on the feedback here. Hopefully you all saw that we launched a series of posts to kick off the project competition:
Suggestion: Rasberry Pi Pico4ML ? From their site: “The single-board microcontroller – powered by Raspberry Pi’s RP2040 chip – to support all Tensorflow Lite Micro tiny machine learning examples with an onboard camera, screen, microphone, and IMU. Run without wiring, or even untethered”
Uctronics Kit for Wake-Word Detection, Gesture Detection and Person Detection
Arducam Pico4ML TinyML Dev Kit: RP2040 Board w/ QVGA Camera, LCD Screen, Onboard Audio, Reset Button & More](https://www.uctronics.com/media/catalog/product/cache/5d1f08909b337dbf64263212361f58e9/b/0/b0302-02.jpg)
Cost: $49.
1 Like
This is now available in Calgary, Canada at Solarbotics.
This is great!
We are thinking of writing a new book on TinyML based on Markdown so that everyone can contribute to it, would be neat if someone is interested in sharing the RasPi version of instructions.
3 Likes
Hi Hal,
Thanks for the info on Solarbotics. Nice resource to have. Thanks and I’m a great admirer of Canada.
Wishing you the best,
Pete
Hi Vijay,
The idea of a new book based on Markdown would be really interesting. As for the RasPi versions of instructions, I’m no expert—just starting out. I do have a few classes on it lined up later and once I have a working demo, I’d be glad to share anything I’ve learned along the way. But my main focus right now is on this class. I’d hate to flunk it focusing on something else. 
Awesome, keep chugging along!
Please let me know how we can help?
Hello everyone.
How is it going?
Is the competition still available?
Thanks
It show sold out yet. I worry about what time we can get one.
Hi Everyone!
Thanks for all the follow-ups on this, my mailbox had conveniently decided that all these emails were spam and wiped up into the spam box. Hence, I hadn’t seen that people were asking and contacting me about this! Sorry about this.
We are extending the competition, as there is a severe hardware shortage. In case you haven’t been aware, there is a serious global shortage of semiconductors.
That aside, there are several people who have spent time and effort on the projects, and we want to recognize the effort. We want to write a blog post that recognizes the effort of everyone, and also sort through the winners (post extension). So stay tuned, more news coming on this soon!
Thanks again for everyone’s patience and effort.
3 Likes
Is the competition still on?
what about a cooperative one? . is just… christmas 