3D scanners are a great way to create models for 3D printing. The problem is that most are either too expensive or too low quality for good 3D printing.
ATLAS 3D solves this problem by using the Raspberry Pi to be both inexpensive and high resolution.
How it works
ATLAS 3D works by illuminating an object with laser light and then using 3D triangulation to generate a point cloud for each location where the laser hits the model. Neighboring points are then connected as triangles to form a 3D model. This model can be used as-is for many purposes or it can be processed in a software package such as Meshlab to make it water-tight and print ready.
All of the software runs onboard the Raspberry Pi, so there are no required drivers or software packages to install. A web browser is used to communicate with the scanner on your home network. Once a scan is performed, the web browser is used to download the resulting models.
In the winter of 2013, I purchased a 3D printer with the goal of being able to print everyday items that I needed around the house. Websites such as thingiverse were a great source for models but often didn't have the models that I needed. I determined that I needed a 3D scanner in order to create my own models from objects that I had. I first discovered the Makerbot Digitizer but it cost more than my 3D printer did. I then began experimenting with 123D catch but I didn't like relying on a cloud based solution which could disappear at any minute, and the model quality was not good enough. Most of the other scanners I discovered were targeted towards commercial metrology and were too expensive for the average consumers.
At this point, I decided to create my own. Similar to the Makerbot Digitizer, I decided to make it a turntable based scanner since a full 360 degree scan can be performed with minimal fuss. I also chose to use lasers instead of photogrammetry in order to generate high resolution models in a relatively small amount of time. When these requirements were established, the Raspberry Pi was an easy choice since it has an excellent 5 megapixel camera add-on, is able to drive lasers and motors, and has enough memory and CPU power to perform a 3D scan.
I quickly got the software into a state where it could produce decent scans and generated a breadboard prototype. I posted some early 3D scan results on the Soliforum 3D printing forum and the positive response convinced me to create this ATLAS 3D Kickstarter so that everyone can create their own 3D scanner.
ATLAS 3D is the product of four hardware iterations and countless software updates of the original prototype.
ATLAS 3D compares favorably to the competition for a quarter of the price.
Open Source and Open Hardware
ATLAS 3D is open source and open hardware. FreeLSS, the software that powers the scanner has been released on Github under the GPL license. The source OpenSCAD files for all of the 3D printed parts are delivered as part of each kit. The electronic design information will be published during week 8 of the project.
ATLAS 3D is provided in the form of kits. No soldering or electrical knowledge is required to build the ATLAS 3D kits. All of the components simply screw together. The user must print the plastic parts for each kit.
The Full Kit
The Full Kit contains everything but the printed parts. This includes the new Raspberry Pi B+, a 5 megapixel CMOS sensor, Wifi adapter, 2 laser modules, electronics board, power supply, stepper motor, rubber coating, rubber feet, the 3D printable files, and all of the fasteners and cables.
The Bring Your Own Pi Kit (BYOP)
The Bring Your Own Pi (BYOP) kit contains everything that the Full Kit has except for Raspberry Pi, camera, and Wifi adapter.This package is great for those who already have a Raspberry Pi B or B+ they would like use it.
The Printed Parts
These parts are not included as part of any kit and must be printed by the backer. The STL and OpenSCAD files will be delivered so that backers can print the parts and make any desired modifications.
Stretch Goal #1: Autoupdate Feature - $120K
If ATLAS 3D reaches $120K in pledges, an autoupdate feature will be developed that allows the user to easily update to the newest version of the scanning software. The scanner will notify the user when an update is available and give them the option to automatically download and apply the update. This feature will be built into every Full Kit and instructions will be provided for BYOP Kit backers to enable this feature.
Stretch Goal #2: Video Assembly Instructions - $150K
If ATLAS 3D reaches $150K in pledges, in addition to assembly documents, video assembly instructions will be provided with each kit. The video instructions will guide the backer step-by-step through the full ATLAS 3D assembly and calibration process.
Stretch Goal #3: Real-time Scan Preview - $200K
If ATLAS 3D reaches $200K in pledges, a software feature will be developed that allows the user to view the 3D model in real-time as the object is being scanned. This would occur directly in the web browser by utilizing WebGL. It would give the user an interactive preview of the scan results and allows them to quickly make decisions such as increasing the scan detail and making lighting adjustments. This feature would be completed in July of 2015 and would be delivered with the autoupdate feature of Stretch Goal #1.
Below is the expected project plan. It begins at the conclusion of the Kickstarter campaign.
- Disruptive Press - http://www.disruptivepress.com/
- Wize 3D - http://www.wize3d.com
- Soliforum - http://www.soliforum.com
- Filastruder - http://www.filastruder.com/
Uriah Liggett is the founder of Murobo LLC. He has a bachelors of science degree in computer science and over 11 years of professional software development. Over the years, he has worked on a wide range of software from 3D graphics engines (CAD) and photogrammetric alignment algorithms to web services and mobile apps. Recently, he has taken an interest in robotics and computer vision.
Risks and challenges
ATLAS 3D, like all laser scanners, has trouble with shiny, black, and furry surfaces. However, most of the risks and challenges have already been overcome from the previous prototype and software iterations. The biggest risk that remains is the availability of quality components at low volume.Learn about accountability on Kickstarter
- (30 days)