I recently blogged about my departure from MBition after a four year stint helping to build a software department within Mercedes Benz. Having contemplated various possible paths, I decided to leave my comfort zone and move away from automotive on-board software, which has been my life for the past 11 years.
Instead, I will help Autoliv Research’s ML/AI team to help them build awesome detection tools to help save more lives. This means working with a group of very smart people ranging from domain experts on things such as psychology, bio-mechanics, machine learning, embedded systems, mechatronics and more. I’m really really excited about this – so much fun to learn.
It is still early days and I still have a lot to learn, so I don’t know the details yet. I’ve been meeting people from all around the world for the past week and a half, and will continue to do so until Christmas, but the real fun starts in 2022.
The team is growing, so if you are interested in joining me in a quest to save lives and learning about real life applications of ML/AI in small embedded systems, don’t hesitate to reach out.
In the autumn of 2017, me and my family moved down to Berlin for a month to start the infotainment software part of MBition. We were a handful of people at Mindspace in Mitte. I’ve never done so many interviews in my life – and I’ve never had the opportunity to get to know so many great people.
Fast forward to today, and we have grown to over 400 people in the project and are about to deliver an amazing product. The architecture team has grown from a one man team consisting of me, into one of the best groups of engineers I’ve ever worked with.
We wanted to change the world. We could not change everything, but I feel that we are bringing improvements to the project, as well as building a better foundation for future work at Mercedes Benz. I very much want to buy the first car with the MBient platform in it, as I know it will be an amazing product.
This has been a fantastic journey and one of the most challenging, but also rewarding, parts in my career. However, all things must come to an end. I’m sad to leave, but I’m also happy to say that I’m staying with the project as a part of the MBition Advisory Council. I’m happy not to go to Berlin twice a month, but I’m already missing the city, so I’m glad that I know that I will come back.
There are so many I’d like to thank and mention so I cannot begin to try to list you all – I’ll just forget someone. You know who you are! Keep up the great work and let’s meet the next time I’m in town!
Continuing my adventures in FreeCAD, I now had to design a top to my Raspberry Pi enclosure. The FreeCAD people were friendly and told me to checkout spreadsheets.
However, since I already started down another path, I decided to name my constraints. Something I picked up from a random YouTube tutorial I cannot seem to find again.
Having done so, I can then refer to the name in a formula and thus, reuse the old value.
I just got myself a 3D printer. It is my second one – I built a reprap back in 2013 – but this one is slightly more convenient to use. I went for the Creality CR6-SE. It works good enough for me out of the box, even though 3D printing is a topic where you can spend infinite time tweaking things.
I used to use OpenSCAD to create my models, as it maps well to my way of thinking (i.e. coding). However, I’ve decided to take some time to learn some new skills and this time around I went for FreeCAD.
FreeCAD is a free, as in freedom, parametric CAD program with a vibrant community. I’m new to CAD – parametric or not – so this is a learning experience for me.
I decided to create a case for my Raspberry Pi and PiTFT based home automation control panel shown in the video below.
To make this happen I first drew the important parts of the Raspberry Pi and the PiTFT hat, to be able to check that the case actually fits the contents. Then I dove into the creation of an enclosure.
The process of sketching, padding, rinse, and repeat is easy and somewhat tedious. I bet there are lots of ways to optimize this. E.g. by parametrize some dimensions to avoid having to “hardcode” them into each sketch. I guess smarter people can share some reference points between sketches too, e.g. to align the holes in the Raspberry Pi to the pegs and peg feet. I’ll learn this the next time around.
Having designed a bottom half, I decided to test fit it, so time to slice it up and then print it. Yes, I use the slicer that came with the printer, as I don’t want to tinker with the printer, but rather with the things I print.
Below, you can see the end result from the first test fitting. Hopefully the picture below shows something that fits as I bravely wrote this while printing. If not, then I really learned something, because I made a mistake.
