154. The 3D Printer That Teaches Itself To Print

The Next Byte Podcast · 26 Dec 2023 · 14m

AI in practice: 2 real uses of AI described

What this episode covers

Daniel and Forbode discuss a collaborative project between MIT, ETH Zurich, and the startup Inkbit that adds a vision-controlled feedback loop to 3D printing. The system uses an inkjet printer with 16,000 nozzles to monitor the print in real-time, allowing for course corrections and the use of materials that smear or don't harden quickly. This innovation enables multimaterial printing of complex geometries like robot hands and heart models at speeds up to 660 times faster than state-of-the-art inkjet printers.

AI in practice

Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.

ManufacturingInkbit660 times faster

An MIT/ETH Zurich/Inkbit team implemented a vision-controlled closed-loop system on an inkjet 3D printer to monitor prints in real-time, allowing the machine to correct errors mid-print and use volatile materials that previously could not be printed.

“This team from Inkbit said that they were able to produce the same parts with the visual control system, about 660 times faster than other state of the art 3D Inkjet printers.”

Speaker 1 · Hear it at 10:59

Manufacturingabout a 2nd

The 3D printer uses a camera to scan each layer after printing, processes the image in about one second, and adjusts specific nozzles (e.g., nozzle 13,582) to drop less material, preventing smears and enabling the use of soft, dynamic materials.

“every time they print, they take a scan, they process it and all that takes about a 2nd. And then you tell each dropper that's within the ink jeopardy. every, there's like. … 16,000 different nozzles. Crazy. But that's the level of control they have.”

Speaker 2 · Hear it at 4:41

Mentioned

Companies
MIT, ETH Zurich, Inkbit, Mouser Electronics, HP, Smile Direct Club, Makina
Products
Formlabs, Stratasys

Other shows on the same ground, closest first.

  • Hackaday Podcast · 1h 16m

    Ep 217: The Unintentional Space and 3D Printing Episode

    Hackaday editors Elliot Williams and Tom Nardi discuss the ESA Jupiter-bound probe's stuck antenna, a research project from the Hasso Plattner Institute that automates laser cutter kerf measurement, and the evolution of the Coco Press chocolate 3D printer into a pre-order kit. They also cover Christopher Helmke’s modular 3D-printed screw production system, tools for cleaning 3D prints, repurposing old smartphones as Linux computers, and the upcoming transition to 2D barcodes.

  • Hackaday Podcast · 44m

    Ep 328: Benchies, Beanies, and Back to the Future

    Hosts Elliot Williams and Kristina Panos review a diverse range of hardware hacks, including Jan Rutz’s ultra-fast 'Minuteman' 3D printer that prints a Benchy in under three minutes using dynamic temperature control and dry ice cooling. They also discuss Chris from Sound Workshop’s 3D-printed woodwind instrument, Truck’s 3D-printed cowbell made from high-temperature PPS-CF filament, and Malcolm Wilson’s infrared photography hack using a Raspberry Pi in a vintage camera body. The episode concludes with a look at Chisanbop finger-counting history and a retrospective on obsolete 1985 technology from 'Back to the Future'.

  • Hackaday Podcast · 1h 16m

    Ep 217: The Unintentional Space and 3D Printing Episode

    Editors Elliot Williams and Tom Nardi discuss the ESA Jupiter-bound probe's jammed radar antenna, a research project from the Hasso Plattner Institute for automated laser cutter kerf measurement, and Christopher Helmke’s modular 3D-printed screw production system. They also cover the Coco Press chocolate 3D printer, tools for 3D printing, repurposing old smartphones as Linux devices, high-altitude ballooning history, and the upcoming transition to 2D barcodes.

  • Additive Snack · 34m

    Latest Advancements in AI for AM Monitoring & Quality Assurance with Interspectral

    Isabelle Hachette, CEO of Interspectral, explains how her company applies AI and data fusion from medical imaging to additive manufacturing process monitoring. She details their software platform, AM Explorer, which ingests multi-sensor data to provide automatic error detection and root cause analysis for metal 3D printing. Hachette discusses how these insights help companies like Volum E reduce inspection time by 90% and accelerate the transition from prototyping to certified production.

  • Hackaday Podcast · 42m

    Ep 368: A Pencil that Draws Against You, 3D Printing Stuff, and Tablet, Shmablet!

    Hackaday editors Elliot Williams and Kristina Panos review recent community hacks, including a laser-powered mist-and-mirrors multi-view display by ancient, a robotic drawing assistant called Penpal from the University of Chicago, and a new Orca Slicer fork enabling supportless 3D printing overhangs. They also discuss a trackball 3D controller by David Liu, a fully mechanical 3D-printed pinball machine by Steven, and a guide to CubeSat design by Francis Drew. The episode concludes with commentary on GPS III satellite interoperability and a debate on the utility of tablets versus smartphones.

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by The Next Byte Podcast. The episode itself belongs to its publisher.