ARMORA turns the repeated jobs at home over to small, vision guided robotic arms that learn you and never stop working. One hub runs them all.
Buyers face two bad choices. Either a giant humanoid that is costly, unproven, and unsettling, or a single-purpose gadget that does one thing and learns nothing.
Large two-legged robots move around the home and feel like a threat, and the price sits far out of reach for almost everyone.
Arms now range from $100 kits to $799 consumer models with cameras and AI. But they are general platforms or DIY, not one job done reliably in a real home.
A dishwasher, a robot vacuum, a pet feeder. Each solves one task in a silo and none of them adapts to the person.
Put the intelligence where the work happens. A fixed arm on a known counter, in a known spot, with a known camera view, faces a far simpler world than a humanoid, so it can hit the reliability a home actually needs. Start with one job done perfectly, then add jobs and arms.
Same spot, same view, one task. The 99-out-of-100 that humanoids struggle to reach.
Every arm sees what it grabs. No blind grippers.
A faulty arm is pulled and a spare clicks in.
The Orb adds jobs over time. One arm is useful, two are better.
The deepest fear about home robots is a machine that walks toward you. ARMORA removes that fear by design. Each arm is fixed where you mount it and reaches only as far as its job.
Light jobs use suction clips you can move in seconds. Heavier jobs use a screw-on base. Power it on and it maps the space and learns the environment on its own.
About 50 cm of usable reach. Holds ~500 g at full stretch, up to ~1.5 kg close to the base. Engineered upward from there.
Suction clips for light jobs, moved in seconds, or a screw-on base for steadier, heavier work.
No setup service. It powers on, maps the room, and learns by itself.
Power stays clean and cable-free: swap a battery cartridge, dock the whole arm to charge, or let it reach a charging station and charge itself. An always-on low-voltage cable is an option for fixed stations. Solar is a later accessory.
A small camera at each arm tip, the size of a car sensor.
The Orb hub, with a mic array. It hears you and runs the fleet.
Orb on wifi, arms paired over Bluetooth. A quick pair, not wiring.
Automotive-style sensors: light, heat, proximity, distance, touch, motion.
Six carbon joints, a swappable gripper, a hot-swap battery.
The same supply chain as today's low cost arms, assembled into one system.
A capable arm is cheap to build, and stationary home arms began shipping in 2026. What no one has won is a task-perfected, reliable, well-priced home product. That gap is opening now.
Home and service robotics: tens of billions of dollars, growing double digits.
~20M affluent smart-home households in the US able to spend $500 to $1,000 on a home helper.
~0.5% of SAM, about 100k units at ~$800 = ~$80M, seeded through Yesodi and affluent early adopters.
Illustrative bottom-up estimate. Pricing and shipping status from a 2025-2026 survey of SO-101, Weave, SwitchBot, 1X, and Dobot.
Arms mount where the work is and pass objects down the line, each doing one step. The Orb keeps the timing.
A fridge arm opens up and lifts out what the eye finds.
It passes the item to the next arm in the line.
Wash, store, or prepare. Each arm does one job.
The Orb remembers the pattern and needs fewer commands.
In 2026 stationary and modular home arms are shipping. That confirms the direction and removes the "is it possible" risk. Nobody has yet won the attainable, task-perfected, reliable home arm. That is the opening.
| Player | Type | Price | Status / limit |
|---|---|---|---|
| 1X (NEO) | Humanoid | $20k or $499/mo | Walks the home; hard tasks still use a remote human expert |
| Figure (03) | Humanoid | Not consumer priced | $1B+ raised at ~$39B; general humanoid, huge complexity |
| Sunday Robotics (Memo) | Stationary, training | Not yet shipping | $165M raised; still gathering data via capture gloves |
| Weave (Isaac 0) | Stationary arm | $3,999 or $249/mo | Folds laundry; 600W mains; still uses teleoperation |
| Weave (Isaac 1) | Mobile, two arms | $7,999 or $449/mo | Powerful, but mobile and costly |
| SwitchBot (onero A1) | Home arm | $799 | ROS2 / MoveIt2; closest price rival; general, not task-perfected |
| Open-source, hobby (SO-101, Dobot) | Kits | $100 to $1,000 | Capable now, but kits, not consumer products |
| ARMORA | Task-perfected home arm | $499 to $999 | One job at industrial reliability, then expands |
Researched August 2026. Weave, Sunday, and SwitchBot validate the stationary and modular thesis, and show the problem is harder than a bare parts cost suggests. The real race is reliability, not whether it can be built.
They must solve walking, balance, navigation, and manipulation. ARMORA fixes the workspace, so it only solves manipulation. Far more reliable, far cheaper, and shipping-focused.
Same stationary thesis, but ARMORA targets a lower price and one perfected job, and puts intelligence where the work happens, without depending on teleoperation.
Those are general or DIY platforms. ARMORA is a task-perfected appliance. The moat is the skill, the reliability data, and the installed locations, not the servos.
| Capability | ARMORA | Humanoids | Other arms | Appliances |
|---|
A low entry model brings buyers in, then trims and options lift the average order. Three revenue lines: the hardware ladder, the arms and options people add over time, and a monthly spare and service plan. Payments via PayPal and PayerURL.
| Starter kit margin (on $799) | ~$527 |
| 1 to 2 extra arms ($399 each) | ~$270 |
| End-effectors + power | ~$70 |
| Service plan (24 mo, net) | ~$300 |
| Estimated value | ~$1,170 |
A new skill is learned with a cloud AI model, at a small cost per household, about 3 dollars. Once learned, a skill can be frozen and run locally on the arm with no further AI cost. Owners can stop the learning once they are happy, so ongoing cost is a choice, not a subscription trap.
Built in China on the same supply chain as today's low cost robot arms. The open-source SO-ARM100 (six bus servos plus boards, 200 dollars retail) is the sanity check.
| 6 bus servos (joints) | $64.92 |
| Gripper servo + mechanism | $12.50 |
| Camera eye (OV5640) | $22.00 |
| Compute (RPi Zero 2 W)* | $15.00 |
| Sensors (ToF, IMU, light, temp, force) | $6.85 |
| Battery 2S + BMS + bay | $9.50 |
| PCBA, shell, wiring, mount | $22.90 |
| Arm BOM total | ~$154 |
| ESP32-S3 module (wifi + BT) | $2.60 |
| Mic array + speaker + amp | $4.30 |
| Custom PCBA | $5.50 |
| Sphere enclosure + LED + USB-C | $8.90 |
| Orb BOM total | ~$21 |
Design lever: move the AI brain into the Orb and put only an ESP32-S3 on the arm to save about 12 dollars per arm. * This model conservatively keeps the compute board in the arm.
Add assembly, tooling, packaging, freight, and duty. The US import tariff is the single biggest swing, so a low and a high tariff case are shown.
| 100 units (first batch) | ~$616 to $684 |
| 1,000 units | ~$272 to $332 |
| 10,000 units | ~$214 to $264 |
| MSRP (Plus) | $799 |
| Gross margin at 1k (low tariff) | ~66% |
At about 37.5 percent, a finished-import arm carries roughly 60 dollars of duty per unit. Importing sub-assemblies and doing final kitting outside China cuts that to about 16 to 22 dollars. The exact HTS classification is settled with a customs broker before pricing is locked.
Researched estimates from real 2025-2026 supplier listings (Feetech, Seeed SO-ARM100, DigiKey, Alibaba, 7pcb, Mekalite, Section 301 schedule). The first 100-unit batch is roughly break-even; margins shown are the 1k to 10k steady state.
ARMORA is a powered robot with a battery, a radio, a camera, and a microphone. The requirements are well defined, but they are not trivial and they are treated as a funded workstream, not an afterthought.
FCC equipment authorization in the US and CE with RED in the EU for the wifi and Bluetooth radios.
UL 3300, the dedicated safety framework for service, communication, and personal-care robots. Force and torque limits and a physical e-stop.
The new EU Machinery Regulation applies from 20 January 2027. Designed to it from the start.
UN38.3 transport and IEC 62133 cell safety, plus compliant packaging and shipping.
Camera and mic fall under GDPR and CCPA. Local processing and no forced app keep exposure low.
RoHS and REACH, product liability insurance, and a settled HTS code with a customs broker.
Certification, safety testing, and tooling are funded together at about 120,000 dollars in this round, run in parallel with EVT and DVT. Compliance is a scheduled milestone, not a footnote.
The unfair advantage: the founder also runs Yesodi, a growing business network of high-value members. ARMORA launches into it, so the first pilots, pre-orders, and affluent early adopters come warm, at near-zero acquisition cost. Video and creators take it wide from there.
Yesodi is a founder-owned network of business owners and high-value professionals. It gives ARMORA a warm first market, the exact affluent buyer for a premium home robot, and a source for the 30 to 50 pilot homes, without paying for reach.
Every use case is a demo. Coffee on wake, the car wash, the handoff line. Built for TikTok, Reels, and Shorts.
Send units to home and tech creators. Real footage beats any ad, and it compounds.
Amazon listing plus an answer-first site, so search and AI assistants surface ARMORA for home robot arm queries.
Year one is pilots and pre-orders, not mass sales. The plan is to prove reliability first, then scale one job at a time. These are conservative illustrative estimates on the researched unit economics.
| P&L (estimate) | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Units (pilots then sales) | 300 | 3,000 | 12,000 |
| Revenue | $0.2M | $2.4M | $10.0M |
| Gross margin | 45% | 58% | 64% |
| Gross profit | $0.1M | $1.4M | $6.4M |
| Operating expense | $0.9M | $1.5M | $3.9M |
| Net (EBITDA) | -$0.8M | -$0.1M | +$2.5M |
Illustrative estimates, not guidance. Year one is a pilot and pre-order year to prove a measured failure rate. Actual results depend on reliability, CAC, tariff treatment, and yield.
Today: the concept, this deck, the product site, and a growing waitlist. This round is not about 20,000 customers. It is about proving one useful job at industrial reliability in real homes, which is what unlocks the larger robotics round.
Build the engineering prototype of ARMORA One on a proven arm base, with a wrist camera and one flagship job: the morning coffee. File IP.
Build 10 working engineering units. Start FCC, UL 3300, and UN38.3. Open pre-orders and take paid deposits.
Place units in 30 to 50 homes, sourced warm from the Yesodi network. Log thousands of autonomous task runs and a measured failure rate. Publish uncut demo video.
Hit the reliability target, convert deposits to sales, add the second job, and raise the larger round to scale production.
The proof point that de-risks the next round: an uncut video of a $799 ARMORA doing one useful household job 100 times across different homes, with the failure rate on paper.
US duty is the biggest cost variable. Mitigation: import sub-assemblies and kit outside China, settle the HTS code with a broker, and hold pricing power at a premium MSRP.
Hardware is hard. Mitigation: build on the proven SO-ARM100 lineage and a contract manufacturer, not new mechanics. Assemble, do not invent.
Large players chase humanoids. Mitigation: own the safe, stationary, home-priced niche they ignore, and move faster.
Servos are half the BOM. Mitigation: dual-source, negotiate below list, and use fewer, cheaper servos on the entry model.
Home robotics is early. Mitigation: a car-style entry model lowers the barrier, and demo-led marketing proves value before purchase.
A robot in the home must feel safe. Mitigation: stationary design, torque caps, e-stop, child lock, and clear certification.
The AI is commoditizing fast. LeRobot, open VLA models, and multi-robot foundation models mean everyone will have "arms coordinated by AI." So the moat cannot be the carbon, the Orb, the camera, or six servos. It has to be the thing rivals cannot copy off a shelf.
ARMORA is founded by Guy Levin, an operator who both builds products and sells them, and who already owns a distribution channel to the exact buyer. This company needs that trio: ship the hardware, run the demo engine, and reach affluent early adopters without paying for them.
Serial entrepreneur with more than 20 years building companies. Founder of the Astrarow Group and of 2410 Games. BA in Business Management, Derby University. A hands-on operator who ships fast.
Scaled GNM Media to about one million shekels in monthly revenue and ran two million shekels a month in ad budgets. Managed teams of up to 700 people. Built more than eight companies, including SortExpress, which holds a PCT patent.
Working with AI since 2017. Ships live products quickly and directs the full stack of a launch.
Also founder of Yesodi, a growing network of high-value business members. A warm launch market for a premium home robot, plus years of paid media at scale.
Sales instructor and a licensed pilot. Comfortable owning outcomes end to end.
Honest status: ARMORA is an early concept. The first hire this round funds is a hardware and manufacturing lead, and a technical co-founder or advisor is actively being sought.
This round does not fund the whole vision. It funds one thing: a task-perfected ARMORA One, proven reliable in real homes. That is enough to earn the larger robotics round. Roughly 18 months of runway.
Seed round, priced or on a SAFE. All financial figures in this deck are illustrative estimates for a concept product, grounded in researched costs, and are to be validated, not treated as guidance.