I’m very proud to announce that my major project at MIT – my work with the Digital Construction Platform – has been showcased in the April issue of Science Robotics. My team couldn’t be more proud to have our work presented in such a high-visibility venue: it’s exciting not just for us, but for how this can help spur further conversation around automated construction.
Our paper is available through Science Robotics, here. You can also read more about the Digital Construction Platform here, and see more images of the dome here. There’s also been a fair amount of popular coverage of our work, which has been really wild to see. Among the better articles covering are work have been:
What this project represents in terms of technical achievement – and as a vision for the future of construction – is something I’m quite proud of. We have assembled a highly specialized, unique robotic system, and used it to validate our spray polyurethane foam printing process at architectural scale. We’ve constructed one of the largest monolithically fabricated 3D printed structures ever built. Our work makes strong arguments about the appropriateness of different kinematic architectures for automated construction tasks, and even more importantly, about how rethinking the way we build through the use of materials like spray foam can dramatically improve the viability of automated construction.
Finally, we’ve had the opportunity to tell a story to the public about how systems like the DCP will change the way we think about construction. I am confident that within the next twenty years, systems like the DCP will be common sites on construction sites throughout the developed world, with more finding applications at natural disaster sites; in Earth’s most inhospitable regions; and even someday on other planets. I am honored to have made some small contribution toward that goal – and I’m excited to see how we get there.
In preparation for graduate school (which starts on Tuesday – yikes!), I purchased a gorgeous new MacBook Pro Retina. I got the deranged idea in my head that I’d like to not only be able to run OS X and Windows on this laptop, but Ubuntu as well, and happily enough found that Travis Llado had come up with the same terrible idea.
I followed his tutorial basically to the letter, installing Yosemite, Windows 10 and Ubuntu 14.04 LTS, but ran into a few snags: however, once you figure out how to get through these, it’s really not such an awful process, and the final product seems pretty stable. I wanted to share the workarounds that I came up with, as well as a few other observations.
- During Linux installation, I did not create a swap partition, in the hopes of reducing the total number of partitions on the machine to avoid Windows freaking out over a MBR drive with more than 4 partitions. I don’t think this is necessary, and there may be downsides to not having a swap partition – please chime in if you have thoughts on this.
- After Step 7 in Travis’s documentation, my computer wouldn’t automatically restart Boot Camp. When I forced it to restart in Boot Camp (Option + Power to boot, and then selecting the EFI Boot disc that showed up), Windows would see the Boot Camp partition as free space, and would refuse to install to it. As near as I can tell, the Linux installation screws up the hybrid MBR that Boot Camp creates, causing Windows to think that there are too many partitions on the disk for it to install safely.
To get around this, I followed Rod Smith’s suggestion here. After following these steps, I was able to Option + Power boot, select the EFI Boot disc for Windows, and successfully install. Rod – who maintains rEFInd and gPart – is an expert on partition management, and I highly recommend reading his site if you want to understand more about what’s going on behind the scenes with rEFInd/partitions.
- My understanding, from what I’ve read on Rod’s site and elsewhere, is that the partitioning scheme that is used in this setup is likely to get ruined by any major OS updates. This may be true, and as with any setup like this, you should always have a regular, reliable backup protocol in place.
However, FWIW, I’ve found this to be pretty stable so far. To get my Windows license (Win8 Pro) converted to Win10, I wound up installing Win10 –> installing Win8 –> upgrading to Win8.1 –> upgrading to Win10 –> wiping computer clean and reinstalling Win10. Throughout this entire process (with the exception of the wipe), the rEFInd boot manager popped up happily, and I was always able to boot back into OS X (I didn’t try Linux).
- Finally, one weird thing I noticed when installing Win10 the second time (the upgrade). I would boot my computer through rEFInd, the Win10 spinning dots would show up, and then the screen would go blank. Nothing. I killed and restarted once or twice, to the same result. Then, after seeing a mention of Win10 incorrectly detecting an external display being connected during installation, I tried removing the power cord and the Thunderbolt to Ethernet adapter I had plugged in while the screen was black. Suddenly, the screen came on, and I was magically partway through the Win10 setup process. My guess is that the computer was interpreting the Thunderbolt cable as a display.
Anyway, if you try this, best of luck – and please comment if you have any feedback, problems, or insights into what’s going on here.
I’m currently in the process of transitioning to a GitHub-hosted Jekyll webpage, which will eventually live at julianleland.io. However, good web design (unsurprisingly) takes time, so for the time being, julianleland.io is redirecting here. As soon as I’ve worked out the initial kinks, I’ll post a link here to the new site (and will eventually transition fully to it) – stay tuned!
Another post with some links to resources I’ve found useful: this time, it’s sources for coefficient of friction data.
If there are any resources you like which I’ve missed, let me know!
(Note: Again, I’m not affiliated with any of these groups, and have no responsibility for/ownership over the content of these pages – I just think they’re useful! Additionally, be aware that a) the data from these sources is unverified – at least by me – and b) friction coefficients are challenging to determine conclusively regardless. They can vary widely based on geometry, surface finish, other surface conditions: you should always verify coefficients experimentally in critical applications.)
I frequently have to generate ISO tolerance zone/fit callouts in my work. I struggled with this for a while, but have (mostly) gotten my head wrapped around the system, and wanted to share some of the resources I’ve found useful for designing these fits:
- Theoretical Machinist ISO Tolerance Zone Calculator: AWESOME tool, gives you a visual display of the tolerance bands of the hole and shaft fits that you’ve specified, showing how much clearance/interference you’re designing in. Doesn’t have every possible fit, but all preferred fits and some less-common ones. This is what I use most of the time when I’m generating tolerance zone/fit callouts.
- MDMetrics Fits and Tolerances PDF: Lots of information on preferred fits (first, second and third choice), with some info on International Tolerance grades and angle tolerancing. Tables of fit ranges are kinda useless once you’ve got the calculator linked to above, but the other info is worthwhile.
- Misumi Tolerancing Cheatsheet: Gives real-world examples of different fit types.
(Note: I’m not affiliated with any of these groups, and have no responsibility for/ownership over the content of these pages – I just think they’re useful!)
Briefly, some exciting news: I’ve been featured by ASME in their My Engineer’s Notebook series! The My Engineer’s Notebook series interviews up-and-coming early career engineers about their interests, their heroes and influences within engineering, and (as you might guess) what’s in their engineer’s notebook. I talked with ASME about the SR-71 Blackbird; my company’s new robot, the Proficio; and David Macaulay’s The Way Things Work. Check it out here.
Some photos of a particularly epic part that we’re building on the 3D printer at work – because 3D printing is just that cool:
MAKE Magazine (which I’ve subscribed to since it first started, and which I’m a huge fan of) has organized a national Day of Making for tomorrow, June 18th, alongside the White House Maker Faire. As part of this program, they’re asking Makers to sign on to a pledge promising to support the development and growth of a vibrant, inclusive Maker community. I’ve signed the pledge, and wanted to make a plug for it here as well: I firmly believe that “bringing the tools to the people” is one of the best ways to promote technological literacy, economic development and help nurture the next generation of engineers, scientists and artists.
Sign the Maker Pledge here!
As an engineer, I do a lot of sketching, and need high-quality pencils. One of my favorite is the Uni Kuru Toga Roulette, which incorporates a small mechanism to automatically rotate the pencil lead so it maintains a point. The mechanism is good for drawing (some people complain about mushiness, but I don’t find it to be a problem), great for writing, and the Roulette is a solid, well-made pencil.
Unfortunately, the Roulette is only available in a 0.5 mm lead thickness, even though other versions (like the Kuru Toga High Grade) are available in 0.3 mm. I decided to try taking apart the Roulette, to see if I could combine 0.3 mm guts with the Roulette body.
There’s only limited information on how to disassemble the Kuru Toga, so I thought I’d post this for future users. Here’s a photo of the Kuru Toga Roulette, fully disassembled:
To take the pencil apart:
– Unscrew the silver tip from the black metal pencil body. This will reveal a black plastic tip, which can be pulled off of the pencil body.
– Grab the plastic section of the pencil body, and pull it away from the metal knurled section. This may take some force/twisting. It will eventually slide off of the polished metal section. The actuator plunger for the pencil (the plastic clicky part, for lack of a better technical term) will remain attached to the plastic section, and a spring will fall out of the section.
– (Hard Part) Unscrew the polished metal section of the pencil from the knurled metal section. This requires some force (since there’s some type of glue/thread lock between the components), and is not easy to do without marring one of the components. I held the knurled section in a collet in a lathe, grabbed the polished section in the tail stock chuck, and then gently twisted them apart by hand.
– Once you’ve done this, the plastic “guts” of the pencil – the lead carrier, the rotation mechanism, and the clutch – should fall out of the knurled barrel easily.
And that’s it! You can now reassemble any combination of Kuru Toga parts you’d like.