New startup ideas · Physical world · Robotics and physical world
startup idea
Partprint
A shared physics fingerprint for every machined part, so factories stop destroying samples to trust them.
Partprint retrofits cheap acoustic, thermal and spindle-current sensors onto CNC machines, presses and furnaces, and turns each part's production signature into a predicted mechanical property card.
- Software subscription
- Enterprise
- $10-100B market
- Network effects, winner takes most
- Global from day one
5/5
venture judge
8
similar startups, last 2 years (18 all-time)
96%
of 4 nearest real companies still alive
yes
8 matching federal grants and programs
Direction supported by government programs and grants
Test it before you build it
$1,850 · 5 weeks · 25 prospects
For $1,850 and five weeks, prove that quality directors at US tier-2 forging and machining suppliers will put $2,500 deposits on a paid parallel trial of fingerprint-predicted part properties, before any sensor is installed or lab is built.
Riskiest assumption · Quality directors at tier-2 suppliers can legally cut destructive-test sample counts under their existing customer sampling plans without an OEM contract change, and will pay five figures for statistical evidence that lets them do it
1Focus group: who and where
Director of Quality or Materials Engineering at a US tier-2 forging or precision machining supplier, 50-500 employees, shipping safety-critical parts (suspension components, fasteners, structural brackets) under customer specs with batch destructive-test sampling plans, currently cutting up 5-20 good parts per lot and eating the scrap cost this quarter
where to find 25 · Forging Industry Association member directory and the Precision Machined Products Association member list (directories); IMTS 2026, September 14-19 at McCormick Place in Chicago (event); local ASQ section meetings and the Practical Machinist forum (communities where these quality and shop leaders already are)
2Sell first, build later
An 8-week parallel trial starting November 2026: Partprint instruments two machines on one line, runs 200 of the supplier's scrap parts through an accredited destructive-test lab, and delivers a report comparing fingerprint-predicted tensile and fatigue values against the lab results, formatted for the supplier's customer quality team
the ask · $9,500 per trial, $2,500 deposit at signature, balance on report delivery
a real yes · A real yes is a countersigned trial agreement with the $2,500 deposit paid and an instrumentation date on the calendar; compliments, unpaid data-sharing offers, and requests to see results from someone else's trial first do not count
3Small experiments
The first one attacks the riskiest assumption; each ends with a number that says whether to run the next.
1. Fifteen sampling-authority calls
$150 · 10 days
Book 15 calls with quality directors pulled from the FIA and PMPA directories and ASQ section contacts. Ask three things with numbers: annual destructive-test spend including scrapped parts, whether their customer specs fix sample counts or reference sampling plans they control, and who signs off on reducing samples. One founder runs all calls off a shared question sheet; log answers in a spreadsheet.
keep going if · 9 of 15 report destructive-test plus scrap spend above $75,000 per year, and 8 of 15 confirm they can reduce sample counts under their own sampling plan without OEM contract changes
2. IMTS meeting sprint
$1,400 · 6 days
One founder attends IMTS 2026 in Chicago for three days with a one-page trial spec, targeting forging and machining suppliers exhibiting in the metal-cutting and quality/metrology halls. Book 12 sit-down meetings in advance from the exhibitor list, pitch the paid parallel trial, and ask for a scoping call the following week.
keep going if · 5 of 12 meetings end with a scoping call on the calendar for a paid parallel trial
3. Sell the paid parallel trial
$300 · 14 days
Run the scoping calls and sell a written 8-week trial: instrument two of the prospect's machines, destructively test 200 of their scrap parts at an accredited third-party lab such as Westmoreland Mechanical Testing or Element, and deliver a prediction-vs-lab report they can show their customer. Price $9,500, invoiced with a $2,500 deposit through a standard invoice and e-signature, no code written until two are signed.
keep going if · 2 signed trials with $2,500 deposits paid, out of 12 scoping conversations
4Collect a deposit up front
Tesla took $1,000 refundable reservations for the Model 3 and $100 for the Cybertruck before building either: the deposit is the measurement, not the revenue.
$2,500
per prospect, refundable
how · A paid design-partner trial invoiced up front: the quality director signs a one-page trial agreement and the supplier pays a $2,500 deposit by ACH against the $9,500 fee, because enterprise-adjacent suppliers pay invoices, not checkout links, and a director can approve $2,500 without procurement set up: Stripe Invoicing ↗
what it reserves · One of three November 2026 trial slots, the founders on site for instrumentation, and 200 destructive tests on the supplier's own alloy and process rather than a generic benchmark
refund · Refunded in full if Partprint misses the agreed start date or the supplier cancels in writing before instrumentation day.
target · 2 paid deposits ($5,000 collected) from 12 scoping conversations within 35 days
Go: build it if
2 trials signed with $2,500 deposits paid, plus 9 of 15 calls confirming annual destructive-test spend above $75,000 and 8 of 15 confirming they control their own sampling plans: build the sensor kit for the first trial
Kill: stop if
0 deposits after 12 scoping conversations, or fewer than 4 of 15 calls confirming they can cut sample counts without OEM signoff: the stated risk is real, contracts pin the testing and the subscription buys no cost out
5 Scripts to run itoutreach message, landing copy, deposit terms · click to open
outreach message
Found your shop in the FIA member directory. If you forge or machine safety-critical parts, you are probably cutting up 5 to 20 good parts every lot to satisfy a sampling plan, and eating the scrap. I am building a way to predict tensile and fatigue results from a machine's own acoustic and spindle signals, verified against an accredited lab on your parts, and I am running paid 8-week parallel trials with three suppliers before building anything else. Do you have 20 minutes this week to see if your specs leave room to cut sample counts?
landing page
Cut destructive-test scrap without touching your customer specs $9,500 buys an 8-week parallel trial: 200 of your scrap parts lab-tested against machine-signal predictions, report formatted for your customer Book a 20-minute scoping call
deposit terms
Your $2,500 deposit reserves one of three November 2026 trial slots: we instrument two of your machines, destructively test 200 of your scrap parts at an accredited lab, and deliver the prediction-vs-lab report. The deposit applies to the $9,500 trial fee. It is refunded in full if we miss the agreed start date or you cancel in writing before instrumentation day.
Would you run this test?
One tap. The yes-share feeds the Demand pillar of this idea's score; nobody sees who answered.
Budgets are out-of-pocket estimates for a team of one to three, US market. Size the deposit to the deal, and check the terms before taking money in a regulated line.
Scorecard
One score that balances how trendy the idea is, the demand for it and its potential for 100x, with competition measured relative to every other idea in the catalog. Recent startup trends first, government priorities second.
76
Idea Score, 0-100 · raw 47.2 x 1.61
Crowded
competition: more crowded than 75% of ideas · headwind x0.63
+4.2
government priorities, secondary (60 matching grants)
Trend
69
Is the wave forming now? 2025-26 entrants vs 2023-24, rounds since 2025, the sector's live-batch direction, the 2026 trend analyst.
- Entrants 2025-26 vs 2023-24 (similar companies)96
- Rounds announced 2025+ in the sector31
- Sector direction (live batch)100
- 2026 trend analyst50
Demand
52
Does anyone want it? YC's current RFS, companies already paid for something similar, the operator judge, founders' yes-rate in decks, readers who would run the test.
- YC asks for it (current RFS: idea / sector)30
- Someone already pays (similar companies, recent / all-time)100
- Operator judge: real pain25
100x potential
90
Can it return a fund? The venture judge (double weight), market-size and moat axes, neighbours still alive, the technologist judge.
- Venture judge100
- Market size axis67
- Moat axis100
- Neighbours still alive75
- Technologist judge100
Score = 100 x cbrt(Trend x Demand x 100x) x (1 - 0.5 x crowding) + government bonus (max 5), calibrated so the 95th-percentile idea scores 90 (order never changes). A geometric mean: a weak pillar cannot be papered over. Percentiles are among the 272 ideas in the catalog; the terms matched were shared, physics, fingerprint, machined, part, factories, stop, retrofits.
The idea in full
- What
- Partprint retrofits cheap acoustic, thermal and spindle-current sensors onto CNC machines, presses and furnaces, and turns each part's production signature into a predicted mechanical property card. To learn the mapping the company builds and operates its own destructive-test reference labs (tensile, fatigue, CT) on three continents, running tens of thousands of parts a year against their fingerprints. Enterprise manufacturers and their tier-2 suppliers subscribe per machine per month to skip batch destructive testing and to demand fingerprints from suppliers in one common format.
- Why now
- The radar shows the manufacturing wedge is being rebuilt right now: twentyfour26, Antropi Robotics, Ultrasonium and ProvenMetal all entered in the F26/S26 cohorts, and cohort volume jumped from 40 in 2023 to 88 in 2024 and 79 already in 2026, while AiSupervision's 2022 attempt at a factory production line OS died as a dashboard with no physical ground truth. NSF has two standing programs, Advanced Manufacturing (AM) and Manufacturing Systems Integration (MSI), plus CMMI, that fund exactly this sensor-to-property translation, and the I-Corps award on metal viscous thread printing (~$50k) shows the same translation path being funded at seed scale.
- Wedge: first customer and entry point
- One automotive tier-1's forged suspension line: instrument twelve machines free, run its scrap parts through the founders' single reference lab, and prove the predicted fatigue curve matches the destructive test on a thousand parts before charging anything.
- Path to 100x
- Global industrial testing and quality inspection spend sits in the $10-100B band, and the fingerprint format is the asset: once an OEM tells 400 suppliers to report Partprint cards, each supplier joins to keep the contract and every joined machine sharpens the priors for the alloy and process it runs. That is a one-format market where the second-best dataset is worthless, so the first company to 100,000 fingerprinted machines takes most of it.
- Ceiling
- If property prediction only ever works within one alloy and one process family, the company becomes a per-vertical testing lab with lab economics rather than a network.
- Closest real companies, as the generator saw them
- Ultrasonium and ProvenMetal make parts faster and cheaper; Partprint does not make parts at all, it prices the confidence in parts other people make. Dawn Industries diagnoses and fixes automation cells from machine telemetry, which is uptime, not part properties; AiSupervision (graveyard, 2022) tried the same telemetry layer with no destructive ground truth to calibrate against.
- Main risk
- Manufacturers keep destructive testing anyway because their customers' contracts specify it, so the subscription buys no cost out.
Five judges
Each judge scores every idea in the catalog with a named rubric; the venture judge decides whether a card is shown at all (4-5 is venture-grade).
Venture investor
5/5
A one-format market in $10-100B testing spend where an OEM mandating cards to 400 suppliers compounds the dataset no rival can rebuild.
Bootstrapper
1/5
Operating destructive-test reference labs on three continents before charging anything is capital this lens will never underwrite.
Operator
2/5
Customer contracts specify destructive testing, so an OEM must force 400 suppliers onto the format before anything is saved.
Technologist
5/5
Owning destructive-test reference labs on three continents to calibrate spindle and acoustic signatures is physical ground truth AiSupervision lacked and nobody clones fast.
Risk
2/5
The network only forms if one automotive tier-1 mandates the card to 400 suppliers, which is key-customer concentration disguised as a format.
trends
3/5
Why-now leans on standing NSF programs rather than a dated shift, but robotics-physical cohort surging to 13% of S26 plus $260M in named grants earns points.
Similar startups in the directory
Companies whose pitch matches most of the idea's terms (shared, physics, fingerprint, machined, part, factories, stop, retrofits): 18 all-time, 8 from the last two years. Same matching as Idea Check.
Metal manufacturing, faster, cheaper, and more versatile.
Moon Hotel -> The Intergalactic Construction Company of Earth
AI Supply Chain Manager for CPG
100% automated inspection for consumer electronics parts
Automatic Diagnosis & Fix for Industrial Automation Cells
Work with agents that already know you
The intelligence layer for simulating and scaling industrial vision systems.
Software tool for designing robots and machines
Quick Algorithm offers AI-powered tools for data analysis and automation.
Turn ideas into physical products
Building the open operating system for the automated physical economy.
Physics-Based AI Digital Twins for Intelligent Manufacturing
The generator's reference companies
Real companies the model named as closest when it wrote the card, with their fate. A check mark is a company the radar could verify in its directory.
Public money in this direction
US federal grants, SBIR/STTR awards and open opportunities from the radar's public-money feed, matched to the idea's terms; the sector totals give the context.
8
grants and programs matching the idea
313
startup-relevant grants in Robotics and physical world
$320M
awarded in the sector, tracked
3
opportunities open now in the sector
- SBIR Phase II: AI-Enabled Acoustic Signal Interpretation for Automated, Standardized Quality Control Testingawardhigh relevance
National Science Foundation · SBIR Phase II · $1M · posted 2026-08-24
- Derivation and Validation of the Heart Failure CardioScore: an AI Diagnostic for HeartFailure using Novel Biomarkers and Clinical Dataawardhigh relevance
NIH / NHLBI · SBIR phase II · $1M · posted 2026-09-01
- EAGER: Leveraging Human-centered Technologies to Enabling Micro-retrofit Energy-saving Opportunities in Homesawardhigh relevance
National Science Foundation · CPS-Cyber-Physical Systems · $150K · posted 2026-08-12
National Science Foundation · I-Corps · $50K · posted 2026-08-10
- I-Corps: Translation Potential of a Digital Pet Health Management Software Platformawardhigh relevance
National Science Foundation · I-Corps · $50K · posted 2026-08-10
- STTR Phase I: Compact, Comfortable Wristband for High-Fidelity Gesture Recognition via Distributed Strain Sensing, Analog Multiplexing, and Edge Machine Learningawardhigh relevance
National Science Foundation · STTR Phase I · $305K · posted 2026-08-05
- SBIR Fast-Track: Automated Creation of Material Models for Multi-Polymer 3D Printingawardhigh relevance
National Science Foundation · SBIR Fast-Track · $2M · posted 2026-07-21
- A Wearable Sensor Platform for Remote Monitoring of Motor Impairment in Huntingtons Disease and Prodromal Huntingtons Diseaseawardhigh relevance
NIH / NCATS · SBIR phase II · $1M · posted 2025-09-01
Market signal
What the radar sees in Robotics and physical world: new companies by cohort year, the forming YC batch, and outcomes since the February snapshot.
Robotics and physical world · 78 → 61 → 207 → 130 → 152 new companies 2022 → 2026 · 89% aliveYC F26 live: 19 in this cluster, 15% of the batch (was 14% in S26)Since February, of 252 YC companies here: 2 acquired, 3 shut down, 36 rewrote their pitch
Design attributes
The card is one cell of a designed set: every axis below was chosen before the text was written, and the text had to realize it.
- Buyer
- Enterprise
- Business model
- Software subscription
- Path to 100x
- Network effects, winner takes most
- Market size
- $10-100B market
- Capital intensity
- Capital-heavy (hardware, bio, infra)
- Speed to revenue
- R&D first, revenue after 3 years
- Technical depth
- Deep tech: ML, hardware, bio
- Go-to-market
- Founder-led sales
- Moat
- Data moat
- Geography
- Global from day one
- Regulation
- Unregulated
- Vibe
- Boring business
Listed under
An idea sits in its own sector and in any sector its text clearly touches.
More ideas like this
Physical world · Robotics and physical world
Sortie
Airspace clearance and audit-grade flight records as one API for drone operators.
Sortie is an API that small drone operators and the software vendors serving them call to get a flight authorized, logged and provable: airspace and waiver checks, operator and aircraft credential validation, geofence rules per jurisdiction, and a tamper-evident flight record that insurers and regulators accept.
Physical world · Robotics and physical world
Dexchange
The certified marketplace where enterprises buy validated robot skills from independent developers.
Dexchange is a marketplace where enterprises running mixed robot fleets buy task policies - machine tending, weld inspection, palletizing - built by independent skill developers.
Physical world · Robotics and physical world
Voltstack
A plug-in home battery that installs itself and joins the grid programs for you.
Voltstack is a 5kWh stackable home battery that plugs into a normal socket, needs no electrician, and handles its own grid paperwork: the unit detects its country and utility from meter signals, files the interconnection or notification form, and enrols the household in whatever demand-response or feed-in scheme it qualifies for.
Physical world · Robotics and physical world
Tactura
Tactile skin modules that give any industrial robot a sense of touch.
Tactura makes force-sensing skin panels and fingertip modules that bolt onto existing industrial arms and grippers, sold as catalog components with a dev kit ordered online and volume pricing after.
Physical world · Robotics and physical world
Crewbay
The marketplace where small businesses hire certified robot crews by the job.
Crewbay is a global marketplace where small businesses - warehouses, farms, cleaning contractors, facility owners - book robot work crews from local operators the way they book equipment rental today.
Physical world · Robotics and physical world
Skillnest
A subscription skill layer that teaches any home robot your house, learned from every other house.
Skillnest is a consumer subscription that rides on top of home robots sold by other manufacturers: the household demonstrates a chore once by phone or by guiding the arm, and Skillnest turns that demonstration into a policy that runs on their machine and, once anonymised and verified, on everyone else's.
Fictional company written 2026-08-23 from MarkosWeb data; the companies, grants and numbers around it are real and tracked. Treat the idea as a research prompt, not a plan.