Episode 401: Matrix Voice Controlled Robot

Just heard watching the new Episode of Matt / airbornesurfer that snips.ai was sold to SONOS in November'19 and therefor this is not a good project to get into anymore.

Thanks for the feedback, Dubbie!

the snips.ai setup was one of the easiest - I dipped my toes in some of them and it definitely was good to start with (tutorials where on point) and the results were great.

7 Servos are indeed excessive and it would need 2-3 servos more so I'm really really happy with it - I want to rotate the head base and be able to pan/tilt/roll the head good thing the SG90ies are so cheap!

I really liked the robot, especially the head part and the facial expression possibilities. This is something I am interested in. But seven micro servos to get this effect does seem quite excessive. The SNIPS.AI looks quite difficult to get going as well, although some other projects I have seen on Element14 that use it have indicated that it is simple to get going. Maybe I'll have a go one day.

Dubbie

Learning is always good.

DAB

Thanks for the feedback!

When I started to revisit this project for the video, I wanted to do a lot of things at the same time.

This video shows - more than my other videos - how I defined the project while designing it.

On other projects I went through the process afterwards and skipped the things I did in parallel that didn't work out.

But still this project and video frustrated me a bit, because it didn't have that "all worked out perfectly fine, like planned" feeling.

I talked a lot at MakerFaire about what's next for the robot, before making the video. My main problem probably

was trying to find a lot of solutions in parallel, because of the nature and shortness of time in making that video.

That way I "wasted" time on problems that weren't important in a later state.

In the end I was successful regarding the project, but I was a bit drained after all that and it shows in the video, I think.

Still, I learned a lot from it and try to improve for future videos..

> FYI, it is always a good idea to sit someone else down and walk through each of these steps with them as you explain what you think you want to do.

This is why I subscribe to the thought process of rubber duck debugging:

https://en.wikipedia.org/wiki/Rubber_duck_debugging

Nice project.

I am always entertained when people think they can just throw a project together and expect it to work.

There is a reason why us engineers spent a lot of time in class learning about the components and more importantly, the process of defining, designing, building and testing a project.

Each step prepares you for the next and forces you to consider things you had not thought of.

FYI, it is always a good idea to sit someone else down and walk through each of these steps with them as you explain what you think you want to do.

You will almost always reach a point where you suddenly realize that you forgot something very important.

Taking time to work through the process always saves you time in the long run. Yes I know a lot of people who say they just sit down and do it, but they are just lucky if they succeed at all.

The process gets you repeatable results.

If you want successful projects, then learn and use the process.

DAB