Most 3D printing news is about the printer. This one is about the step before the printer, and it is the step most people skip. Washington State University researchers just published an electronic skin for prosthetic hands that senses pressure and temperature at roughly ten times the resolution of commercial glove sensors. The part that matters for anyone who needs a custom-fit part, though, is how they made it: they scanned the prosthetic first, mapped the design onto that scan, and only then printed it. WSU calls it "scan-model-print." We call it Tuesday.
What WSU actually built
The work, led by graduate student Hongyi Shen and Professor Kaiyan Qiu in WSU's School of Mechanical and Materials Engineering, was published in Cell Reports Physical Science and announced by the university on August 20, 2026. The problem they set out to solve is one anybody who has tried to fit a flat sensor onto a curved object already knows. Existing electronic skins are expensive, cover small areas, and get worse at sensing the more you bend them to fit a custom shape.
Their fix was to stop fighting the shape. According to the WSU release, a scanner captures the geometry of the prosthetic, software maps the sensor layout directly onto that scan data, and the structural layers are printed on an SLA resin printer to match the freeform surface. Laser-cut flexible electrical layers are then assembled into thin sandwich modules that hold both a pressure-sensing matrix and temperature sensors. The modules snap together without adhesives, so a damaged tile can be swapped instead of replacing the whole skin. A neural network calibrates each sensing element and strips out manufacturing noise so the data is usable in real time.
The team has filed a provisional patent disclosure and is now working on an actuator that would turn the sensing signals into stimulation an amputee could feel.
I have an oddly shaped object and I need a part that fits it exactly. Can someone 3D scan it and print the part?
Yes. A 3D scan captures the real geometry of your object, we model the mating part directly against that scan, and then we print it, which is the same scan-model-print workflow WSU used to fit sensors to a prosthetic hand. This is how we handle the jobs that never fit a catalog: a bracket for a discontinued machine, a cover for a hand-formed housing, a mount that has to sit flush on a curved surface, or a replacement for a part that only exists as a worn original. Measuring a freeform surface with calipers is guesswork. Scanning it is data.
The WSU team was not printing something and hoping it would conform; they were designing to a surface they had already captured. That is the difference between a part that "mostly fits" and one that seats on the first try. If you have something like this on your bench, our 3D scanning and reverse engineering service is built for exactly that.
What the modular, snap-together design means for your parts
The second idea in this research is one we push on engineering buyers constantly: build the assembly out of small, replaceable pieces instead of one big fragile part. WSU's sensor modules snap together like building blocks. If one fails, you replace one. If you need more coverage, you add tiles. No adhesives, no full redesign.
In our shop that translates into a few practical habits. Design fixtures and covers as tiled or segmented parts when the total footprint is large, so a single damaged section is a small reprint instead of a full rebuild. Use snap-fit features where the material allows it (we keep ABS snap-clip strain under about 1.5 percent), and use press-fit crush ribs where you need a firm hold without hardware. And print the mating face against a scan, not a sketch, whenever the surface is anything other than flat.
For anyone shipping a product line, modular design also means low inventory and shipping a replacement tile instead of a full unit. That is how we run build-and-ship production runs with no minimum order.
My printed part looks right but never seats flush on the curved surface. What am I doing wrong?
You are almost certainly modeling the mating face from measurements instead of from a scan, and small errors on a curved surface add up to a part that rocks or gaps. Calipers give you a handful of points. A curved or organic surface needs thousands. Scan the surface, bring the mesh into CAD, and model the contact face as an offset of the real geometry. Then leave a small clearance for first-layer squish, orient the part so the contact face prints cleanly, and test-fit a small gauge section before committing to the full part.
Why a Rockville shop cares about a Pullman lab
We do not build electronic skin. What we build every week is the middle of this story: parts that must match an existing object's geometry, in materials that hold up. WSU used SLA resin for fine detail. For parts that need heat or chemical resistance we move to engineering polymers, including PEEK, ULTEM, and PPSU on our high-temperature printing service. The lesson from Pullman is simple: the fit problem is solved before the print starts, and the scanner is what solves it.
Brands: we review tools that make scan-to-print easier
If you make handheld 3D scanners, structured-light systems, scan-to-CAD software, or flexible and conductive filaments, we test gear in real production and write about what actually works on the shop floor. Reach out at info@dcadditivepros.com if you would like us to evaluate a product.
Frequently asked questions
I need a replacement part for a discontinued product and I only have the worn original. Can you make one?
Yes. We 3D scan the original, rebuild it as a clean CAD model, correct the wear, and print a new one in a material suited to the job. One piece or a hundred, no minimum order.
How accurate is 3D scanning for making a part that has to fit?
Accurate enough for most mechanical fits. Handheld and structured-light scanners capture curved and freeform surfaces far better than hand measurement, and we verify critical dimensions with calipers before we cut the final model.
Can a 3D printed part be designed to snap together without glue or screws?
Yes. Snap-fit clips and press-fit crush ribs are standard features we design into printed parts, with the geometry tuned to the material so the fit holds without cracking.
What file do I need to send to get a custom part 3D printed?
A STEP or STL file is ideal, but you do not need a file at all. Send us the physical object, a sketch, or photos with dimensions and we can scan or model it for you.
Sources: Washington State University press release, "Researchers develop electronic skin for prosthetics to sense temperature and pressure," August 20, 2026; 3DPrint.com, "3D Printing News Briefs, September 5, 2026." The research paper appears in Cell Reports Physical Science.