opserv

[opserv]

3d model test

Bracket Orientations2 by haifa3d on Sketchfab

[opserv]

It seems even casinos are looking at #automation these days. In 2015, the Tokyo police raided an illegal casino and found a robot worker in what was the first known case of automated gambling dealer. In fact, the robot turned out to be a robotic arm, rather than a humanoid.

An article in Casino.com from 2016 says robot dealers are 30% faster and far more accurate than their human counterparts. The article also muses that rookie players may prefer a robot dealer to avoid humiliation when making silly mistakes.

I recently came across the DoBot M1. A sub-$2000 robot that approaches light-industry standards, with decent mechanics which allows it to move quickly and accurately, and pretty decent software (which will inevitably improve over time)

A robot like this can be easily fitted to, among other things, deal cards - and replace a human dealer remotely serving players and earning 

Even the gimmicky 3D printing application of the robot is somewhat useful, when you consider that by putting the robot on an (optional) rail, you can get a build volume over 5 times the average 3D printer. But lacking a proper build platform and other amenities, it is limited to the niche of extra-large, rough models. The optional 4th axis add-on, if properly utilized, can allow it to perform more complex additive manufacturing tasks like embedding objects in the print while printing.

And get this - the robot is designed and made by chinese startup. One with a proven record of delivering on 2 previous KickStarter campaigns.

see the humble DoBot M1 in action - 

[opserv]

View this post on Instagram

A post shared by @fablabtom

Spent an hour trying to make one of those Fidget spinners. It came out a bit too large to spin comfortably. About 6 min carve on the Carvey from 8 mm birch. I'll have another go this week. @inventables #carvey #inventables #fidgetspinner #cncmilling

[opserv]

Just passed by this car parking in Haifa #diy

[opserv]

Educational device. CNC milled birch wood. Designed with onshape. Used for demonstrating the bernoulli effect. Unfinished prototype, design by Javier verba, madatech science museum. Should be open sourced shortly.

[opserv]

Beautiful project, cnc milled birch on the fablab shopbot. The design was copied from a Kayak seat. Design and 3d cam on Fusion 360, by Daniel Kachaev and mishael Kraus. Milling time was only 1 hour thanks to adaptive strategy of Fusion 360.

[opserv]

Purim Holiday, using a graphtec plotter and 240g paper stock. Design by yulia Agronov

[opserv]

New typo

[opserv]

Tinkercad puzzle workshop

[opserv]

Creative students

[opserv]

Accessible AI

After listening to the wonderful Siraj Raval explain how back propagation works. Wonder what happens when we let neural nets optimize for the best neural net structure to perform a certain task? The next step towards general intelligence?

[opserv]

In interaction design you want to be as versatile as possible with your haptic interactions. touch is great - but touch sensor modules are cumbersome and hard to fit into our physical environment. luckily you can hook up an external electrode to a metal pad, or even a metal object like a lamp. here I am using a large metal paper clip. 

[opserv]

Finally! someone had the guts to do what I never dared to do! 

PrintrBot has a CNC named CrawlBot. It is using regular aluminium profs instead of V , but using a V wheel.

I thought they printed a special wheel, but not so. You can see from the marks on the wheel where the wear is on a single point on the diameter.

PrintrBot were never hesitant to compromise for “good enough” solutions - they pioneered using nylon fishing line to save on timing belts. The CNC machine is being sold for at least 2 years without problems. This means

V-profiles are expensive and hard to ship. Possibly a better wheel can be designed to improve on this.

[opserv]

Naked chef in the FabLab, or - How to start something big

I discovered the Jamie Oliver "Food Revolution" project via TED. This is a guy who seems genuine in his convictions - to change food culture, school by school , restaurant by restaurant. Ideology and reality show hype aside, his approach is very relevant to anyone starting a successful FabLab or community makerspace. If the "food revolution" is about eating better and being healthier, the "maker/fab revolution" is about making things and gaining self reliance. Both are about empowerment through gaining skills and knowledge. In both cases people are a bit reluctant to engage at first. And neither is sustainable without the cooperation of a community.

These are some of Jamie's approaches I am implementing in the Fab Lab:

Stay low but look up - Jamie's strategy is double-pronged - he starts at the bottom, in one kitchen, in one school. At the same time he starts to move things at the top - making alliances with senior members of the community, hospital officials, government and so on. At every milestone, he creates a connection between the two - a challenge for the team at the bottom, rewarded by acknowledgement by the guys at the top, who get a chance to visit and feel involved. The same applies for a new fablab - Make sure you get the proper attention from the right people, but spend enough time developing real projects.

take care of your kernel - When Jamie sets up shop in a new place, instead of trying to engage everyone at once, he starts with a small group of 5 to 10 people, who get some group training and close attention. He then leverages this group to gain wider traction in the organization. I believe that fablabs must not be a service provider to the community, but a platform for people in the community to lead projects. Iv'e seen places that lean to the former - If they do well financially, they become TechShop-like, and if not, the manager of the lab/space quits their job due to exhaustion and lack of satisfaction. The kernels are intended to lead projects and activities for their social and professional circles. And by doing that, they serve as your ambassadors. more on that in another post.

get out of the nest - Jamie opens his equivalent of a fablab - a community kitchen with free cooking lessons, in the middle of the town. But most of his efforts are in outreach activities. I believe that if you want to really make a change, expand your audience beyond the "usual suspects". Get out there, in the street, in the schools, in social activities in your community (not just maker faires!). I totally accept that some people flourish in a familiar and welcoming environment. If you have such people in your team, great - let them take care of the nest; the awesome routine of people taking first steps in the maker world.

Regarding my current situation , I can say that not being able to give proper attention to my "kernels", is my biggest worry out of the three. I lost precious kernels in past projects - people with energy, willingness to take responsibility, great skills and/or plenty of free time (precious commodity!)

[opserv]

ככה זה נראה כשאנחנו בנינו את המייקרבוט בסביבות פסח

[opserv]

אתמול אמיר , שי וחנן באו לבקר ופיקחו על תהליך פיתוח הרפלקטור לאנטנה של WIFI

[opserv]

**How to Fit Large Objects on a 3D Printer Plate** I've been having problems figuring out if objects I design will print properly or not. Recently I found out a method in SolidWorks to fit larger objects and make sure they are not over the limit. 1. Draw a 90x90mm square on a clean sketch and drag and rotate it so it fits around your object. If you are brave , you can increase this as far as your printer goes.. (**never use 100x100mm** if you don't turn off the Perimeter option in SF) 2. In the middle of the square, create a new Coordinate System and align it with your square. 3. Choose Save As and pick STL. click Options, under Output Coordinate System choose the name of the coordinate system you created before. 4. the object will import into ReplicatorG perfectly scaled and aligned!