Hardware Startups funded by Y Combinator (YC) 2026

Flock Safety provides the first public safety operating system that empowers private communities and law enforcement to work together to eliminate crime. We are committed to protecting human privacy and mitigating bias in policing with the development of best-in-class technology rooted in ethical design, which unites civilians and public servants in pursuit of a safer, more equitable society. Our Safety-as-a-Service approach includes affordable devices powered by LTE and solar that can be installed anywhere. Our technology detects and captures objective details, decodes evidence in real-time and delivers investigative leads into the hands of those who matter. While safety is a serious business, we are a supportive team that is optimizing the remote experience to create strong and fun relationships even when we are physically apart. Our flock of hard-working employees thrive in a positive and inclusive environment, where a bias towards action is rewarded. Flock Safety is headquartered in Atlanta and operates nationwide. We have raised $150M in our Series E led by Tiger Global at a $3.5B valuation.

Shepherd Robotics builds general purpose robots for skilled physical work across AI infrastructure and advanced manufacturing. Our robots are powered by robot foundation models and use modular, flexible hardware designed to adapt to different environments and tasks. We train our systems on a combination of real world and simulation data so they can learn complex manipulation skills and generalize across workflows. We are starting with tasks such as data center construction and operations, cable and wire manufacturing, wire harness assembly, kitting, and factory packaging, where traditional automation struggles with variability, dexterity, and changing environments.

Frontier is first to grow biological brains at scale, serving biological neural nets with substrate-level colocation of memory and compute, enabling highly cost effective training of large parameter count models. While academia has traditionally been limited to ~1M neurons in vitro, Frontier has built a pathway to scale to 100M+ neurons by circumventing traditional scaling challenges in biological culture, demonstrating this strategy's functionality by training a biological brain using this scaling primitive to play Frogger. Frontier is now growing exascale compute clusters to produce substrates capable of long-horizon language reasoning while operating at much higher sample efficiencies than traditional architectures, and collocating memory with compute to enable significantly larger parameter counts than would be tractable for training within traditional compute clusters.

Edgerun builds military exoskeletons. The first product is a 10-pound hip system that fits over existing kit and doubles a soldier’s range. Since April 2026, Edgerun built over a dozen iterations of their product through continuous field tests with Ukrainian fighters and secured $150M+ in LOIs. Founder David Venegas previously co-founded Durin (autonomous mining robotics) and worked at Stoke Space (at 19) and Hadrian. He is a high-school dropout, state-championship runner, adventure athlete, and pilot. Edgerun is a Y Combinator company backed by General H.R. McMaster, Cyan Banister, Joshua Steinman and Jason Miller. edgerun.com

We build physical agent harnesses to enable AI to debug, validate, and ship functional hardware and software.

antimattr builds a voice-first hardware interface for human-agent interaction that includes a ring and a pair of earphones. Designed to make more of your voice, the ring can record and summarise your conversations, capture thoughts on the go, or control your agents. The earphones use proprietary voice isolation technology to block out background noise entirely, so you can take a clear call from a loud pub or dictate with real accuracy. Reserve early access at https://www.antimattr.one/project-mnemosyne [use code YC at checkout]

Dream makes AI cameras that catch damage teams miss. Car dealerships, equipment rental companies, fleets, and more move expensive assets in and out all day, and undocumented damage is a costly leak. Dream cameras are simple to set up and work automatically. Each vehicle or machine that passes is photographed from every angle, identified, and reconstructed into a complete condition record. On return or intake, Dream instantly flags new dents, scrapes, and changes your team might miss, giving one source of truth for every asset's condition over time. Ryland worked at X/xAI and built aerial damage-detection vision at Vexcel Imaging. Alexander worked at AWS and trained edge-camera models deployed in the wild across Africa and India to detect poachers.

Physical AI — robots and world models — is bottlenecked on one thing: diverse, real-world data at massive scale showing how humans actually do work in the physical world. It was never written down, and you can't buy it, because no one has ever captured it. Only real local experts, the people who do this work every day, can produce data this authentic and valuable. DeepReach builds the network that reaches them. We design and manufacture wearable stereo capture devices, then place them with local data partners who run collection as their own business — in warehouses, workshops, farms, kitchens and repair shops. They own the local relationships; we supply the hardware, the software platform, the quality pipeline and the payments. Anyone can run a data business. That structure is what produces diversity. Instead of thousands of hours inside one environment, we deliver hours spread across thousands of environments — which is what actually makes robot policies generalize. 475 devices in the field and 100+ data partners are live across 7 countries, adding new partners every week. They've collected nearly 150,000 clips in the past three months, with volume roughly doubling month over month. Our data is in production with multiple frontier model and robotics companies, and we're scaling to 10,000+ devices across 50 countries.

Every serious robotics lab is bottlenecked on the same thing: demonstration data. The field has converged on egocentric video and hand pose as the substrate, and both are now cheap to collect at scale. What neither of them contains is force. A video of someone picking up an egg and a video of someone picking up a rock are nearly identical in pixel space and nearly identical in joint angles. They differ almost entirely in a quantity nobody is recording. Policies trained without it learn geometry and inherit no sense of touch. Hemlock closes that gap at the source. Our device, the Bracelet, is a wrist-worn surface EMG band that reads the electrical activity of the forearm muscles that actually drive the hand, which means it recovers contraction force directly rather than inferring it from motion. This distinction is the whole thesis: force cannot be reconstructed downstream of kinematics, because a hand squeezing hard isometrically and a hand resting can occupy the identical pose. Muscle signal is the only non-invasive channel that separates them.

The technologies that define this century, including data centers, rockets, satellites, and robots, are built from complex metal parts. As these systems advance, the parts get harder to make. Past a certain complexity, casting and machining stop working, and the advanced methods that remain are slow, expensive, and hazardous. Ultrasonium has invented a new manufacturing method to make these critical parts 15x faster and 75% cheaper than the methods they replace, in the U.S. and at scale. Solid metal feedstock goes in and a finished part safely comes out, controlled by a physical AI layer and novel control processes. The systems that matter most, in aerospace, energy, and compute, are demanding metal geometries that can't be machined or cast and are too expensive and slow to make with incumbent methods, at a moment when domestic manufacturing capacity is a national priority. Our founding team has built and innovated some of the most complex engineering systems known to man: quantum computers, superconductors, nuclear reactors, and novel aerospace alloy processing methods.

IMPACT builds autonomous interceptor drones that ship in sealed containers. The box sits at the site it protects, data centers, power grids, airports, bases, with no crew on the pad. When sensors cue a hostile drone, an interceptor pierces its own frangible lid, launches, and rams the target, guided by onboard vision that keeps working when GPS is jammed. Legacy effectors cost $1M+ per shot against $30k threats; ours cost less than the threat and scale with the swarm, the same class of interceptor already downing Shahed drones over Ukraine. Sites subscribe to protection instead of buying missiles, hardware, reloads and monitoring included: Air Defense as a Service. 180M+ USD in signed letters of intent.

CarSignal is the AI-native operating system for independent auto repair shops. Most shops still run each repair across disconnected management software, scan tools, parts catalogs, service information, texts, spreadsheets, and knowledge kept in technicians’ heads. That fragmentation slows diagnosis, creates duplicate processes, and leaves customers waiting. CarSignal connects the full repair process in one system. AI agents research vehicle-specific failure patterns and surface the tests that confirm or rule out each cause. The same system carries each job from booking and intake through inspections, estimates, parts, approvals, payment, and follow-up, so technicians spend less time searching and updating systems and more time fixing cars.

Stoa provides pricing and trading infrastructure for AI hardware, live with an RFQ marketplace where companies buy and sell GPUs through vetted dealers real-time.

Kara is building diamond semiconductors. We begin with growing ultra-high purity diamond wafers for advanced hardware. Used in data centers, spacecrafts, RF systems, and quantum hardware, our diamond moves heat up to 5× better than copper, helping critical systems run cooler, faster, and more reliably. Diamond has spent centuries as a symbol. We are turning it into infrastructure.

Lamb Labs is building MPUs (Model Processing Units), custom chips that hardcode LLM weights directly into the chip. Unlike GPUs, which move weights between memory and compute, MPUs keep the model on-chip, targeting up to 20,000+ tok/s and a 63× higher intelligence per watt.

Applied Electrodynamics builds imaging systems that see through solid matter. Our team previously co-invented the iPhone touchscreen and Nest learning thermostat. We’re starting in construction, helping teams map structural elements and utilities inside walls and ceilings. The same platform extends to security screening and cargo inspection. Our mission is to make the hidden physical world machine-readable across the entire electromagnetic spectrum.

Egoist Machines is building infrastructure for a world where your personal context belongs to you, rather than to the apps and AI companies you use. Today, every assistant and consumer app builds its own version of you. We think that architecture is backwards. Your context should live with you, persist across products, and be available to any model, app or agent you choose. AI Passport is the first layer of that infrastructure. It brings the context you generate across AI assistants and everyday apps into one portable profile you control. Tell ChatGPT something and Claude can already know it, for example. Switch to a new product and it can understand you from the first interaction, without the new company needing to own the underlying data. Apps integrate through Sign in with AI Passport, requesting only the context they need for a specific purpose and duration. Users can see what is being accessed and revoke that access later. Underneath, we have built our own memory infrastructure for fast, compact per-user retrieval and permissioned access. We are distributing AI Passport directly to consumers through first-party products, while integrating it into consumer applications with existing distribution. Every new product that uses Passport gives users another reason to maintain their context there, and every additional Passport holder makes the context layer more useful to developers. The next step is to build a true Egoist Machine: an intelligent system whose memory, incentives and actions are governed by the individual it represents. We call it Digital Domovoy, the hardware extension of AI Passport. Instead of your most personal context ultimately living on somebody else's servers, it lives and is processed locally on hardware you control. AI Passport becomes the permission layer between that private local context and the cloud models and internet services you choose to use. Our North Star is the Egoist Machine: an intelligent system whose memory, incentives and actions are governed by the individual it represents.

The US Warehouse industry spends $75bn on labor each year. Yet, only 10% of US warehouses have any form of automation, because existing robotics are too expensive and hard to integrate. Manifold deploys autonomous robots that handle picking, loading / unloading, and palletizing — priced by pick. Integration in hours. No upfront costs. No retrofitting. No need to rebuild the operation around the robot. Our mission is to make it easy for any warehouse to adopt robotics.

Hardware Intelligence makes chip design tools that cut simulation and debugging time by an order of magnitude. When a chip test fails, engineers need a recording of every internal signal - a waveform. We’ve developed WaveZip, which captures waveforms 500x smaller with 200x less overhead, and Wave, which beats frontier models on industry chip debugging benchmarks. We’re Rishov Sarkar (Georgia Tech Ph.D. in chip simulation) and Athreya Anand (ex-Google, M.S. ML/AI), YC S26. https://hardwareintelligence.ai

Singularity is answering humanity’s call to engineer counter-weapon systems at scale. We build to protect people on the battlefield and at home.

ProvenMetal manufactures the fastest circuit boards in the United States. Domestic circuit board manufacturing is slow and expensive, and while reshoring is happening, the supply chain isn't there yet. We automate the bottlenecks - part procurement, DFM review, and assembly - delivering circuit boards in days instead of weeks.

Robocurve builds open-source tools and independent benchmarks to measure how well robots can do real-world jobs. Instead of relying on unverified demo videos from frontier labs, we score their models on reproducible benchmarks that anyone can trust. Today there are no well-run, standardized robotics benchmarks. Labs evaluate in-house, and no independent group has stepped in to run continuous benchmarking as a service. The result is that no one actually knows how good anyone else is, or where the real frontier sits. Good benchmarks require operating and maintaining physical hardware and real-world setups. Simulation only goes so far, since models that look strong in sim can show large performance gaps once deployed in the real world. Building real-world benchmarks means coordinating job-domain experts, evals engineering, and hands-on robotics all at once. Frontier labs are targeting general-purpose robotics by 2028, yet the field of robotics evals barely exists. Whoever builds the trusted measure of robot capability becomes the reference everyone relies on. We combine backgrounds in AI evals and robotics to build and grow this field as fast as possible. We already shipped v1 of our open-source framework, Inspect Robots, and ran our first pilot scoring a frontier model on a real robot.

Risklytics is the insurance company for frontier tech such as physical AI, robotics, and data centers. Companies answer 6 questions and we bind a tailored policy that other insurance companies won’t.

We manufacture Nori: A sub-$2000 bimanual mobile manipulator that can clean rooms, restock shelves, and fold clothes. Nori can handle basic tasks out of the box, and its software lets anyone teach Nori to automate work specific to their home or business. Nori aims to solve the robotics data bottleneck. By deploying affordable robots to customers around the world, we can gather the diverse data required to train the next generation of generalist robotic policies. Every user of Nori makes Nori smarter. Since launching 6 weeks ago we have already deployed our first robot and have over $440k+ in sales. We can bring general-purpose robotics to homes and businesses that could never afford it before. There are 132.7 million households and 8.4 million employer business locations in the U.S. At $1,688 per robot, selling just one Nori to 0.5% of them represents a $1.2B market opportunity in the U.S. alone.

We build LEGION, a drone swarm that identifies, geolocates, and tracks every emitter on the battlefield using the enemy’s own radio signals. LEGION enables our warfighters to precisely track previously obscured targets like jammers, command posts, and drone operators so they can proactively respond to threats instead of reacting to them.

Simantic is modernizing the embedded firmware industry with hardware simulation developed for the adoption of AI in hardware engineering. Simantic gives both humans and AI the visibility into hardware previously limited by the physical world.

Arlo Industries is building a network of sensors that precisely tracks stealthy drones and aerial threats better than radars. This capability does not exist yet because traditional radars are built on past war doctrines designed to protect single, centralised assets. Modern warfare, however, demands wide-area, persistent coverage that is both passive and economically scalable. By utilising a mesh architecture, Arlo Industries unlocks unprecedented asymmetric economics. The network is fundamentally built to scale: the cost of adding sensors grows linearly, but the tracking accuracy increases exponentially, delivering highly resilient, persistent coverage for a fraction of the cost of legacy systems. Deo founded Arlo Industries after spending over 6 years in the Israeli defence ecosystem, where he lived through multiple conflicts and experienced the Iron Dome and other systems in action against ICBMs and the infamous Shahed drones. Learning exactly how these defence systems could be improved after a Shahed exploded close to his apartment, he built Arlo Industries around the core philosophy that conflict should be concise & precise. The people behind Arlo1 includes a PhD in drones and autonomy and military experts. Arlo Industries has actively showcased this technology directly to Ukrainian militaries on the frontlines, as well as to operators across Europe and the USA.

America does not just need more welders. It needs welding knowledge that can scale. The American Welding Society projects 320,500 new welding professionals will be needed by 2029, with about 80,000 jobs to be filled annually from 2025 to 2029. More than 157,000 current welding professionals are approaching retirement. Welders are only a small share of the skilled-trades workforce, but welding accounts for an outsized share of projected capability demand. The market calls it a labor shortage. AMR sees a learning bottleneck. We are building American-made robotic welding cells that combine computer vision to provide real-time seam tracking and post-weld inspection to turn welding from a scarce manual process into a closed-loop manufacturing system that learns from every weld as it happens.

Tenet Industries is building AI-automated factories for manufacturing of PCBAs. We're using AI and robotics to build a factory in München for instant-quote, fast-turnaround PCBA manufacturing and electronic manufacturing services.

E1O (E1 Optics) is building camera-free AI glasses with a private display and a personal assistant. Follow your coding agents, catch up on email, and bring useful information into view. Connected to the E1O app on your phone. Join the early-access waitlist at e1o.com.

Up to 80% of the most advance AI compute chips sits idle because of slow networking. Piris Labs builds photonic based networking solutions to enable fast, scalable, and low cost AI workload execution infrastructure.

The best grid connection is no grid connection. Voxel builds vertically integrated energy generation and storage systems for data centers that bypass utility connection delays. By leveraging the vast supply of second-life EV batteries and our novel DC microgrid architecture, Voxel generates datacenter power on site and drastically reduces system cost and time-to-power.

We engineer inexpensive, autonomous strike drones designed to operate in the most contested combat environments on Earth! By combining true hardware modularity with an agentic control system that lowers cognitive load on operators, we enable faster upgrades, broader adoption, and more than 2x reduction in lifecycle cost. Our customers and partners include the UK Royal Marine Commandos, several other units in the UK, and 4 other NATO forces, and we are currently shipping prototypes for iterative feedback. Think of us as Anduril, but with worse weather, and better banter.

Robotic factories that build robotics in the US. Currently operating a 12,000 sqft factory in the Bay Area and actively manufacturing customer orders for some of the largest robotics companies in the US and the International Space Station. We’re a group of Berkeley graduate robotics researchers who previously won a full scale autonomous IndyCar competition at 160mph.

Tornyol builds micro-drones that kill mosquitoes. We use smartphone microphones, car park assist sensors and some clever DSP and control to transform 40-gram toy drones into mosquito killers. Because these drones are so cheap and fast, we will lower the cost of mosquito control by 100x which will enable the eradication of mosquitoes from cities, saving more than a million people each year from Malaria, West Nile, Dengue, etc.

Norra helps nursing facilities manage equipment operations. Every year, ~$80 billion are spent on medical equipment like beds, concentrators, and wheelchairs. Nursing facilities continue to operate with zero visibility which leads to $15 billion in equipment waste annually. We use physical trackers and AI agents to eliminate this.

We build brains for robots. We work with manufacturers like Unitree to make robots listen, speak, and act. You can program your robot out of the box by literally talking to it, no coding involved. Robots running Lightberry are emotionally intelligent, always on, and fully autonomous. Just like in Star Wars!

Icarus is building the modern-day U-2 spy plane. Our autonomous solar aircraft fly at 60,000 feet for weeks at a time, eventually months and years, above weather and below satellites. They deliver persistent intelligence and connectivity for the military at a fraction of the cost of existing systems. The stratosphere is the forgotten frontier.

Our agents run hardware engineering tasks that normally take hours or days in just a few minutes, allowing engineers to process 100x more data and information. Our platform allows hardware engineers to orchestrate agents for tasks such as analyzing failures across entire test campaigns, pulling the right data and KPIs from their databases, checking how conditions (like temperature or load) affect performance and reliability, and then document findings in reports and dashboards.

Testing and certifying the hardware of the future

Everyone pays someone to handle the parts of their life that don't need them personally. The cleaner, the bookkeeper, the lawn service. Nobody could handle what lives in your phone. Until Blue. Blue is a small USB-C dongle and iPhone app. Plug it in and say what needs doing. The follow-up you haven't sent, the appointment you haven't booked, the reply you owe. Blue opens the right app, does the work, and tells you when it's done. You don't touch your phone. Works with every app already on your iPhone. The team brings complementary experience at scale: Omar Abdelaziz led Google Assistant’s transition to LLMs and holds one of earliest patents in LLM-based computer control; Peter Krogh spent decades designing voice-first dialog systems and was most recently a design lead at Google DeepMind building actions for Gemini Live; Muhammad Farhan Hossain helped ship Apple Vision Pro, is named on dozens of hardware/systems patents, and has led products from prototype to mass production. We’ve built systems used by billions. Now we’re on to defining the next era of computing.

OS3 is building the platform that makes intelligent robots accessible to every business. Today, robotics is constrained by two fundamental bottlenecks: hardware is too expensive to deploy at scale, and robots are difficult to train because high-quality robotics data is scarce and costly to collect. We built OS3 to solve both. Our two-person team designed and a US-made mobile humanoid robot for a fraction of the cost of commercial alternatives. On the intelligence side, instead of relying on expensive robot-generated training data, we trained our own video-action model on 150,000 hours of real-world video spanning more than 2,000 tasks. Every deployed robot feeds new experience into our reinforcement learning pipeline, self improving robot fleet over time. This creates a robot that can be deployed across industries to automate repetitive, physical work without requiring a new robot or training pipeline for every task. Our robots are already deployed in hotels, with customers now expanding the platform into biotech, hospitals, and data centers, including a wet lab at Stanford. Our mission is to do for robotics what Apple did for personal computing: take technology that was once expensive, specialized, and inaccessible and make it powerful and usable at massive scale. We’re building toward a world with billions of intelligent robots, and OS3 is the platform that makes that possible.

SigmanticAI is building an AI-native hardware development assistant that automates the entire RTL design flow, from natural language to synthesizable HDL and testbenches, inside a seamless VSCode fork. Powered by fine-tuned Verilog LLMs, reinforcement learning, and real compiler feedback, SigmanticAI iteratively refines code until it compiles and passes synthesis, no matter how complex the design. It’s like the Cursor for HDL design. In addition to code generation, SigmanticAI can generate token-level annotations and onboarding documents to accelerate ramp-up for new engineers. With support for both cloud and on-prem deployment, it integrates with existing EDA tools to dramatically reduce debug time, improve collaboration, and accelerate hardware design across teams

GD&T takes years to master. Engineers must encode their design intent into a dense syntax that manufacturers must decode, and not everyone gets it right. Put ten engineers in front of the same drawing and you may get ten different interpretations. Which means errors often surface later in interface stack-ups, manufacturing, or in the field - when they’re far more expensive to fix. That’s why we built F4 - the platform that analyzes GD&T to determine whether your designs are ready for production. Drop in your drawing and 3D file. F4 analyzes every element of your GD&T to assess whether your design is clear, manufacturable, and production ready. To make the drawing easier to understand, engineers can visualize their datums, reference frames, and tolerance zones directly on the 3D model. F4 also calculates tolerance stack-ups instantly - directly from the drawing. What once took hours of manual work now takes seconds. When issues are found, F4 flags them and guides engineers toward resolving the problem. After analyzing your drawing, F4 also automatically generates design and inspection reports - stack-up tables, FAI reports - ready to share with your team. And it does all of this locally on-device with no AI used for analysis. Producing deterministic results engineers can trust.

Building AI agents for autonomous embedded software development.

Flywheel AI converts any existing excavators for contractors to enable remote ops to increase safety and productivity, and use robotics context dataset to train autonomous policies.

Boost Robotics is building autonomous mobile manipulation robots to remotely perform manual tasks in data centers. The surge in AI applications is pushing data centers to scale faster than they can manage, leading to suboptimal operations and staffing shortages amid rising power and compute demands. We save data centers millions in insurance premiums and SLA violations by increasing the frequency of inspection and maintenance while reducing the time it takes to triage and address failures.

NODI is building distributed energy infrastructure to rewire the grid. We deploy and aggregate batteries at the edge and orchestrate them as distributed nodes, providing locational flexibility to the rigid grid — relieving congestion and unlocking capacity at constrained locations, so data centers can connect in weeks instead of waiting decades for reinforcement.

Photonium is building software to automate optical system design. We supercharge optics experts with intelligent tooling to reduce costs and deliver faster. We handle the full design stack — from optimization, verification, sourcing, to prototyping — for AR/VR, quantum, biotech, metrology/chip fab, LiDAR, and more.

What if every experiment you wanted to run could be modeled using AI? Godela is an AI-powered physics engine that gives engineers faster, cheaper replacement to simulations and physical prototypes. No set up, no wait. Just ask—and get simulation quality result.