Tend.ai Proposes Cloud Robotics for 3D Print Farm Automation

 3D printing has numerous benefits, from the ability to customize products to fabricating geometrically complex parts; however, the technology is still limited in terms of output, making traditional production techniques much more appealing for mass manufacturing. Farms of 3D printers are a potential solution to batch production, with some companies relying on numerous machines to 3D print a large group of parts at once.

While they may increase output, print farms may also be labor intensive to operate. To demonstrate the overall utility of its cloud robotics platform, an Oregon startup called Tend.ai has deployed a robotic arm, artificial intelligence and cloud computing to create an automated print farm.

The Tend.ai platform combines machine learning and cloud computing for machine automation. (Image courtesy of Tend.ai.)

Tend.ai has built what it suggests is a universal machine automation solution that combines a number of components into a cloud robotics platform. While the specific make and model of each piece of hardware is left up to the user, all one needs to get started with Tend.ai is a robotic arm, a webcam and a robotic gripper. These tools combined with a preconfigured, low-cost computer from Tend.ai allows for an automated method to handle a variety of machines, such as a room full of 3D printers.

The Tend.ai platform is able to autonomously manage multiple 3D printers at once. (Image courtesy of Tend.ai.)

As demonstrated in the video below, the robotic arm with attached webcam and gripper is trained through the Tend.ai computer module to perform all of the tasks necessary for operating a 3D printer farm. The webcam captures the data displayed on a printer’s LED screen, allowing its human user to make decisions remotely through the cloud. The arm can also be trained to perform such functions as pressing buttons on a printer, as well as removing a print from the printbed and dropping it onto a conveyor belt.

The founders of Tend.ai—Mark Silliman, Robert Kieffer and James Gentes—all have extensive tech experience, with all three leading companies that were eventually acquired. Kieffer worked as a developer at Facebook and Google, while Gentes was employed at software security company Symantec. Silliman said that he has “involved in open-source robotics for years while founding various software startups.”

The idea for the platform sprang from witnessing a friend run a print farm that fabricates cookie cutters for Etsy. Rather than physically monitor and run the 3D printers, the Tend.ai team thought that she could rely on a robot to do most of the manual labor for her. Unlike other print farm management solutions, Tend.ai does not require the machines to be networked together, relying instead on the intelligence of the Tend.ai software as manifested in the robotic arm.

The Tend.ai platform can be controlled remotely. (Image courtesy of Tend.ai.)

According to Silliman, the Tend.ai software builds up its knowledge as it works with customers, eventually resulting in a machine that can even solve problems on the fly. “[T]he ‘brain’ of your system needs to fully understand machines and have the ability to troubleshoot. For example: If filament is left on the printbed after unloading, our system can identify this and goes into a ‘clean mode’ and starts removing the filament,” Silliman said.

The Tend.ai team proposes its platform as an equally viable solution for large, industrial machines and desktop 3D printer farms alike. Given the many issues that can occur with desktop 3D printing—clogged extruders, adhesion difficulties and random flukes—it may be difficult to imagine a robot capable of navigating every variable in a 3D printing environment. However, if one has learned to optimize the various 3D printing parameters to the extent that he or she can operate a print farm, all that may be missing is a robotic arm for performing the tedious tasks of pressing “print” and removing objects from the printbed.

In addition to the cloud robotics system itself, the company hopes to establish a repository for automation processes through an online community that is similar to Thingiverse for 3D printable designs. Through a beta program launching this July, Tend.ai will begin testing its platform with makerspaces and print farms before eventually releasing the final product in 2018.