WEBVTT
00:00:00.000 --> 00:00:02.879
Aptronic Apollo vs One X NEO.
00:00:03.279 --> 00:00:06.480
Which 2026 humanoid has the more scalable design?
00:00:06.639 --> 00:00:09.759
By Akari Sato, Consumer Robotics Specialist.
00:00:09.919 --> 00:00:13.199
Status checked October 27, 2026.
00:00:13.439 --> 00:00:17.600
A humanoid robot can look ready for work and still be far from scalable.
00:00:17.839 --> 00:00:27.839
The real test is whether it can do useful jobs safely, reliably, at a cost buyers can defend, and whether the company can build, support, and improve a large fleet.
00:00:28.000 --> 00:00:32.320
Eptronix Apollo 2 and OneX's NEO are aimed at different markets.
00:00:32.560 --> 00:00:36.560
Apollo is being developed for factories, warehouses, and other businesses.
00:00:36.799 --> 00:00:40.320
NEO is designed for homes and is already listed for pre-order.
00:00:40.560 --> 00:00:49.439
That makes this comparison less like choosing between two versions of the same product and more like comparing two plans for getting humanoid robots into everyday use.
00:00:49.679 --> 00:00:54.399
The short verdict, Apollo 2, has the stronger design for scaling industrial work.
00:00:54.479 --> 00:01:01.520
Its wheeled and bipedal options, swappable batteries, fleet tools, and manufacturing partnership suit business operations.
00:01:01.759 --> 00:01:11.359
NEO has the clearer path to a mass market product, a published price, a lighter and softer body, and a company-run factory with ambitious capacity targets.
00:01:11.599 --> 00:01:18.000
But those targets are not the same as customer deliveries, and NEO's early service still includes remote human help.
00:01:18.159 --> 00:01:20.079
One distinction matters throughout.
00:01:20.319 --> 00:01:23.920
Apollo 2 is Apptronix's current learning and pilot platform.
00:01:24.159 --> 00:01:31.359
Aptronix says its next commercial platform will follow, with real production versions expected in 2027 and beyond.
00:01:31.599 --> 00:01:39.920
NEO, by contrast, is being offered for home pre-order, but as of October 27, OneX had not announced customer deliveries.
00:01:40.079 --> 00:01:45.040
At a glance for the full table, please open this article on robotcomparisons.com.
00:01:45.280 --> 00:01:50.799
The comparison is uneven because Aptronic has not published a full Apollo 2 specification sheet.
00:01:50.959 --> 00:02:00.400
NIO's numbers are easier to find, but many are company-listed specifications, not independent proof of work speed, safety, or reliability in customer homes.
00:02:00.640 --> 00:02:05.200
Two designs built around different work, Apollo puts choice of mobility first.
00:02:05.439 --> 00:02:08.639
Apollo 2 comes in wheeled and bipedal configurations.
00:02:08.800 --> 00:02:10.719
That is a practical design choice.
00:02:10.960 --> 00:02:15.919
On flat factory floors, wheels can avoid the extra balance and fall risk of walking.
00:02:16.159 --> 00:02:22.400
Legs are more useful where the robot must cross spaces built around people or handle routes that wheels cannot manage.
00:02:22.639 --> 00:02:28.879
Aptronix says the wheeled version is designed to fit with existing safety standards for industrial mobile robots.
00:02:29.120 --> 00:02:35.840
That is a design aim, not public proof that a particular Apollo 2 configuration has a safety certification.
00:02:36.000 --> 00:02:42.080
The bipedal version offers more flexibility, but also brings more demanding balance and safety questions.
00:02:42.400 --> 00:02:47.599
This modular approach makes Apollo's design adaptable to a business's site and task.
00:02:47.759 --> 00:02:55.360
It may also let Aptronic collect useful data on multiple body types without building a separate robot platform for every job.
00:02:55.520 --> 00:03:03.120
But buyers still need current dimensions, surface details, and performance data for the exact configuration they are considering.
00:03:03.360 --> 00:03:05.120
NEO puts home use first.
00:03:05.280 --> 00:03:06.960
NEO makes a different trade.
00:03:07.120 --> 00:03:10.639
It is a single bipedal form designed to live around people.
00:03:10.879 --> 00:03:22.000
Its 66-pound weight, soft outer body, washable suit, and quiet operation are sensible choices for a home, where people, pets, furniture, and clutter share the same space.
00:03:22.240 --> 00:03:26.560
Its design is more consumer-friendly than a typical industrial machine.
00:03:26.719 --> 00:03:30.080
But a home is not automatically an easy place for a robot.
00:03:30.240 --> 00:03:36.960
It has narrow paths, fragile objects, changing light, children, pets, and many tasks that are hard to describe in advance.
00:03:37.199 --> 00:03:44.800
NEO's lighter body may make it easier to live with, but does not by itself prove the robot can safely complete chores without supervision.
00:03:45.120 --> 00:03:47.439
Hardware, payload, and working time.
00:03:47.599 --> 00:03:48.560
Size and weight.
00:03:48.719 --> 00:03:54.800
NEO publishes the Clearer Case 1X lists NEO at 5 feet 6 inches and 66 pounds.
00:03:54.960 --> 00:04:04.560
Aptronic's older Apollo materials list a robot at about 5 feet 8 inches and 160 pounds, but Aptronic has not confirmed those figures for Apollo 2.
00:04:04.800 --> 00:04:09.919
Buyers should not treat the older body's weight or size as a current Apollo 2 specification.
00:04:10.159 --> 00:04:12.400
The difference matters for more than shipping.
00:04:12.639 --> 00:04:16.480
A heavy robot may need more careful handling during setup and service.
00:04:16.720 --> 00:04:21.759
A lighter robot is easier to move, but it may have different limits on strength and stability.
00:04:21.920 --> 00:04:26.800
Without current Apollo 2 measurements, buyers cannot make a fair apples to apples comparison.
00:04:27.040 --> 00:04:27.839
Actuators.
00:04:28.079 --> 00:04:29.839
Apollo emphasizes efficiency.
00:04:30.000 --> 00:04:32.079
Neo emphasizes gentle contact.
00:04:32.319 --> 00:04:34.639
Actuators are the parts that make a robot move.
00:04:34.879 --> 00:04:43.120
Aptronix says Apollo 2 uses its own patented actuators, designed for 90% plus energy efficiency, simple maintenance, and mass production.
00:04:43.279 --> 00:04:47.360
The company also says the design reduces reliance on single source suppliers.
00:04:47.600 --> 00:04:54.160
These are important scale goals, but they are company claims, not published fleet energy or repair results.
00:04:54.480 --> 00:04:56.480
NEO uses tendon drives.
00:04:56.639 --> 00:05:03.360
Motors pull tendons to move the joints, more like muscles pulling on limbs than a rigid motor sitting at every joint.
00:05:03.680 --> 00:05:12.560
One X says this design has low inertia and is backdrivable, meaning an external push can move a joint rather than meeting a hard, unyielding stop.
00:05:12.720 --> 00:05:16.079
Its product page lists 95% backdrivability.
00:05:16.319 --> 00:05:20.000
That approach fits a home robot where gentle contact matters.
00:05:20.240 --> 00:05:22.319
The strategic trade-off is clear.
00:05:22.560 --> 00:05:26.879
Apollo is framed around efficient industrial movement and maintainable hardware.
00:05:27.199 --> 00:05:31.279
NEO is framed around softer contact and close work with people.
00:05:31.439 --> 00:05:39.839
Neither has yet published the kind of long-term service data, such as real-world failure rates, that would settle which system is cheaper to keep running.
00:05:40.079 --> 00:05:40.720
Payload.
00:05:40.800 --> 00:05:44.160
Compare the useful load, not the biggest number NEO lists.
00:05:44.319 --> 00:05:50.000
154 pounds lift, 55 pounds carry, and 18 pounds arm payload.
00:05:50.240 --> 00:05:56.240
These figures do not mean that NEO can safely carry a 154-pound object around a home.
00:05:56.639 --> 00:06:01.360
For buyers, the arm payload and carry figures are more useful than the broad lift claim.
00:06:01.600 --> 00:06:10.399
Aptronic's older Apollo materials say the earlier robot could lift 55 pounds, but that is not a current Apollo 2 payload rating.
00:06:10.639 --> 00:06:13.680
Aptronic has not published one for the current version.
00:06:13.839 --> 00:06:22.639
Until it does, procurement teams should ask for limits by reach, speed, object shape, and whether the robot is walking or using a wheeled base.
00:06:23.040 --> 00:06:23.680
Runtime.
00:06:23.839 --> 00:06:25.759
Apollo has a clearer shift work plan.
00:06:25.920 --> 00:06:33.600
NEO lists a 4-hour runtime from an 842-watt-hour battery and says it can plug itself in when it needs charging.
00:06:33.839 --> 00:06:36.240
Its page also lists quick charging.
00:06:36.399 --> 00:06:43.120
These are useful specifications, but buyers still need to test how runtime changes with different tasks and loads.
00:06:43.360 --> 00:06:52.480
Aptronic describes swappable batteries, opportunity charging, and tethering as ways to keep Apollo working towards 7-day, 22-hour operation.
00:06:52.639 --> 00:06:59.360
That is an uptime design strategy, not a published guarantee that the robot will work reliably for 22 hours a day.
00:06:59.680 --> 00:07:04.240
Aptronic does not list a current Apollo 2 runtime or battery swap time.
00:07:04.480 --> 00:07:09.199
For a factory buyer, Apollo's battery plan is the more mature fit on paper.
00:07:09.360 --> 00:07:16.079
For Neo, self-charging is appealing in a home, but the robot still needs to navigate to its charger and dock successfully.
00:07:16.319 --> 00:07:20.399
Hands, sensing, and AI Neo has more published hand detail.
00:07:20.560 --> 00:07:24.319
The Neo products page lists 22 degrees of freedom per hand.
00:07:24.480 --> 00:07:34.480
In a July 2026 update, 1X described its new hand as having 25 degrees of freedom total, 22 in the fingers and palm, plus 3 at the wrist.
00:07:34.639 --> 00:07:42.720
That appears to be a difference in what is counted, not necessarily a conflict, but buyers should confirm which hand revision is included in the NEO they are offered.
00:07:42.959 --> 00:07:48.399
1X says the hands include force and touch sensing and are designed to bend under contact.
00:07:48.560 --> 00:07:57.839
That is promising for tasks like handling small or fragile objects, but a highly articulated hand only helps if the robot can use it reliably and quickly.
00:07:58.079 --> 00:08:01.439
1X has not published customer home success rates for chores.
00:08:01.839 --> 00:08:10.000
Apptronic calls Apollo 2's object handling dexterous, but does not publish a standard hand model or a full set of hand specifications for buyers.
00:08:10.240 --> 00:08:16.560
Google DeepMind has tested Apollo 2 with different hands, including the SharpaWave and Inspire hands.
00:08:16.720 --> 00:08:20.959
That shows Apollo can be used as a platform for AI and hardware experiments.
00:08:21.199 --> 00:08:25.920
It does not tell buyers which hand configuration will ship for a specific deployment.
00:08:26.160 --> 00:08:37.200
NEO publishes more of its sensors One X lists NIO with two stereo fisheye cameras, four microphones, Wi-Fi, Bluetooth, and 5G.
00:08:37.440 --> 00:08:42.240
It also lists an NVIDIA Jetson Thor computer inside its Neo Cortex system.
00:08:42.399 --> 00:08:48.080
This gives buyers a clearer view of the sensors and onboard compute in the advertised product.
00:08:48.320 --> 00:08:55.440
Apptronic's Apollo 2 materials describe perception systems, but do not provide a comparable current sensor list.
00:08:55.679 --> 00:09:04.480
That is an information gap for buyers who need to check camera coverage, sensor redundancy, data handling, or fit with site security rules.
00:09:04.799 --> 00:09:08.080
Both are learning systems, not finished all-purpose workers.
00:09:08.240 --> 00:09:15.759
Apptronic connects Apollo 2, its Artemis control system, Fleet Connect Fleet Software, and its Google DeepMind partnership.
00:09:15.919 --> 00:09:24.720
Robot Park, a nearly 90,000 square foot site, lets Apollo robots collect task data in logistics, manufacturing, and retail settings.
00:09:24.960 --> 00:09:28.639
Apptronic says that data helps train Gemini Robotics models.
00:09:28.960 --> 00:09:31.679
This is a strong learning and operations strategy.
00:09:31.919 --> 00:09:35.840
Robots work, data improves the models, and the fleet can be updated.
00:09:36.080 --> 00:09:41.600
Okay, Google's July 2026 Gemini Robotics 2 results show both promise and limits.
00:09:41.759 --> 00:09:44.960
In its tests, Apollo 2's success varied by task.
00:09:45.120 --> 00:09:54.559
For example, performance on whole body pickup tasks ranged from 45.7% for items on the floor to 76.3% for items on a shelf.
00:09:54.720 --> 00:09:58.000
On some multifinger hand tasks, results were uneven.
00:09:58.240 --> 00:10:06.639
Google reported 92% success for unscrewing a bulb, but 36% for screwing one in, and 40% for closing a zip bag.
00:10:06.879 --> 00:10:15.919
These are controlled tests, not factory shift results, but they are a useful reminder that a successful demonstration is not the same as reliable everyday performance.
00:10:16.240 --> 00:10:20.960
NEO's AI stack is described through Redwood AI and OneX's world model.
00:10:21.279 --> 00:10:26.080
The product is advertised as having basic autonomy and learning more skills over time.
00:10:26.240 --> 00:10:33.039
But NIO also offers scheduled expert mode, where a 1x worker remotely guides it through tasks it does not yet know.
00:10:33.200 --> 00:10:39.039
That makes NEO's early model a blend of autonomy and human help, not a fully independent housekeeper.
00:10:39.360 --> 00:10:43.759
Safety, privacy, and maintenance Apollo is built around controlled work zones.
00:10:44.000 --> 00:10:46.960
Apollo's safety plan includes hardware-level safety zones.
00:10:47.200 --> 00:10:58.879
Aptronix says its impact zone can pause movement when an object is detected within a defined radius, while a configurable perimeter zone changes how the robot responds to nearby people and objects.
00:10:59.039 --> 00:11:04.320
That is useful for structured workplaces where safety boundaries can be set and monitored.
00:11:04.480 --> 00:11:10.240
For buyers, the next step is to ask for the safety case for the exact task and configuration.
00:11:10.480 --> 00:11:16.720
What happens when a person enters the work area, a load shifts, or the robot loses balance or communication?
00:11:16.960 --> 00:11:22.639
The public product page does not provide independent test results or a full list of certifications.
00:11:22.960 --> 00:11:26.879
NEO is softer, but also brings sensors into private rooms.
00:11:27.039 --> 00:11:34.080
NEO's soft lattice body, covered joints, and low inertia tendon drives are meant to reduce injury risk during contact.
00:11:34.240 --> 00:11:39.200
1X also lists a head injury metric and says the body has no pinch points.
00:11:39.360 --> 00:11:45.360
These are design features, not proof that every home task is safe around children, pets, or vulnerable people.
00:11:45.600 --> 00:11:52.879
1X's own guidance says people should remain attentive and follow its safety instructions, especially around those groups.
00:11:53.200 --> 00:12:00.080
NEO's remote help raises a second safety question: who can see the home through the robot's cameras and when.
00:12:00.799 --> 00:12:03.919
1X presents expert support as scheduled.
00:12:04.080 --> 00:12:10.879
The company's FAQ, as reported in October 2026, says owners must accept a session.
00:12:11.120 --> 00:12:21.279
Even with that control, buyers should ask what sensor data is sent, whether sessions are recorded, how long data is kept, and how data is used to train models.
00:12:21.600 --> 00:12:25.279
Public information does not answer all of those questions clearly.
00:12:25.519 --> 00:12:27.759
Apollo brings a different data risk.
00:12:27.919 --> 00:12:36.480
Its data comes from workplaces and customer sites, where video or task records may expose production methods, layouts, or company information.
00:12:36.720 --> 00:12:43.200
Enterprise buyers should settle data ownership, access, retention, and model training rights in the contract.
00:12:43.360 --> 00:12:52.799
Maintenance is a scale test for both Apollo's swappable batteries and Apptronix claims of simpler, more maintainable actuators are built around reducing downtime.
00:12:53.039 --> 00:13:04.799
Jabil's role adds manufacturing and supply chain expertise, but public materials do not give buyers comparable service intervals, repair costs, or field failure rates for Apollo 2.
00:13:05.120 --> 00:13:12.159
NEO's washable outer suit and OneX's in-house manufacturing may simplify some repairs and replacement parts.
00:13:12.399 --> 00:13:17.600
The company also says its hands have been tested for millions of interaction cycles.
00:13:17.840 --> 00:13:27.039
That is a useful engineering signal, but buyers still need to know how often parts need service, who will do that work, and what it will cost after the warranty.
00:13:27.360 --> 00:13:28.879
How they fit the work.
00:13:29.120 --> 00:13:33.200
Factory and warehouse, Apollo is the stronger industrial choice today.
00:13:33.360 --> 00:13:36.799
Its wield base suits repeatable routes on level floors.
00:13:36.879 --> 00:13:41.679
Its bipedal version is the more flexible option for spaces designed around people.
00:13:41.919 --> 00:13:47.200
Swappable batteries and fleet management tools make the product direction fit business operations.
00:13:47.600 --> 00:13:52.799
Apollo says Apollo 2 is used at robot park and at partner or customer sites.
00:13:53.039 --> 00:13:58.639
Reuters reported that Apollo had built hundreds of Apollo 2 robots by June 2026.
00:13:59.039 --> 00:14:02.240
Well noting the company did not share how many were deployed.
00:14:02.480 --> 00:14:09.360
That is evidence of real platform activity, but it is not the same as proven production scale customer use.
00:14:09.679 --> 00:14:15.600
NEO is being built in a factory, but that does not yet show it can do factory work for a customer.
00:14:15.759 --> 00:14:22.639
1X says some NEOs are helping with internal factory tasks, including basic warehousing and logistics.
00:14:22.879 --> 00:14:25.200
Those activities are useful tests.
00:14:25.360 --> 00:14:30.159
They are not yet evidence of large, repeatable deployments at outside companies.
00:14:30.480 --> 00:14:38.480
Retail Apollo has the clearer business case for back-of-house retail tasks, such as moving stock or replenishing shelves.
00:14:38.720 --> 00:14:45.360
Aptronic lists retail as a target area and says robot park robots learn across retail tasks.
00:14:45.519 --> 00:14:53.440
But public sources do not show sustained retail results, such as cases handled per hour or how often a person must intervene.
00:14:53.759 --> 00:14:59.679
Neo's smaller, softer, quieter design may be more comfortable in a people-filled space.
00:14:59.840 --> 00:15:08.799
Yet its home first interface and early reliance on expert support make it hard to recommend for a retail operation without a dedicated pilot.
00:15:09.039 --> 00:15:14.080
Home Neo is the obvious fit by design, but not yet the proven home worker.
00:15:14.240 --> 00:15:17.759
It's quieter, lighter, and designed around home chores.
00:15:17.919 --> 00:15:24.320
Its voice interface, app, self-charging feature, and softbody all make sense for domestic use.
00:15:24.480 --> 00:15:35.039
However, OneX' product page describes basic autonomy for early owners, and the company had not announced customer deliveries as of October 27, 2026.
00:15:35.360 --> 00:15:40.559
Apollo could eventually serve homes, but it is not the consumer product in this comparison.
00:15:40.799 --> 00:15:49.440
Its current strengths and business model are directed at enterprise use, and there is no public home price or consumer ordering channel.
00:15:49.759 --> 00:15:53.519
Price, manufacturing, and the 2027 Outlook.
00:15:53.759 --> 00:15:55.600
NEO has a posted price.
00:15:55.919 --> 00:15:58.159
Apollo is an enterprise quote.
00:15:58.399 --> 00:16:06.159
One X lists NEO at$20,000 for early access ownership or$499 per month for a subscription.
00:16:06.320 --> 00:16:16.000
Its order page lists a$200 deposit and says US deliveries start in 2026, with early access customers first and subscriptions later.
00:16:16.240 --> 00:16:24.159
At$499 a month,$36 months would total$17,964 before any other costs.
00:16:24.399 --> 00:16:28.960
The terms and support package still matter more than that simple calculation.
00:16:29.360 --> 00:16:34.399
As of October 27th, NEO remained in the pre-order stage in public updates.
00:16:34.559 --> 00:16:44.399
1X said its early units were being used for internal development and home testing, and an October 2026 report found no announced customer delivery.
00:16:44.639 --> 00:16:53.600
The company still aims to begin deliveries in 2026, but a buyer should treat a specific arrival date as unconfirmed until 1X gives one.
00:16:53.919 --> 00:16:57.039
Aptronic does not publish an Apollo 2 price.
00:16:57.279 --> 00:17:00.720
Its site directs interested businesses to contact the company.
00:17:00.960 --> 00:17:05.599
Buyers should expect a custom enterprise process and ask for the full cost of a pilot.
00:17:05.839 --> 00:17:11.359
Robot, software, integration, service, training, safety work, and downtime.
00:17:11.680 --> 00:17:16.319
No public price estimate is dependable enough to use as a purchase benchmark.
00:17:16.559 --> 00:17:19.599
Manufacturing plans are not the same as output.
00:17:27.759 --> 00:17:34.240
Jable's global production network could help Atronics scale, but the companies have not announced a public Apollo production target.
00:17:34.559 --> 00:17:44.000
Aptronic also has a research partnership with Google DeepMind, while Mercedes Benz and GXO are among the customer or partner environments named for Apollo work.
00:17:44.400 --> 00:17:47.039
One X is taking a more vertically integrated route.
00:17:47.200 --> 00:18:02.240
It says its 58,000 square foot Hayward factory makes key parts in-house and has capacity for 10,000 NEOs per year, with a target of 100,000 plus units per year by the end of 2027 as automation and additional factory capacity grow.
00:18:02.480 --> 00:18:07.440
The company also says it sold out its first year of production capacity in five days.
00:18:07.599 --> 00:18:16.640
Those are major ambitions, but capacity and pre-order demand do not prove that 100,000 robots will be built, shipped, and supported in 2027.
00:18:16.960 --> 00:18:19.839
The near-term timeline is different for each company.
00:18:20.160 --> 00:18:21.039
Aptronic.
00:18:21.200 --> 00:18:23.440
Continue pilots through 2026.
00:18:23.680 --> 00:18:29.119
The company says Apollo 2 data is feeding the next commercial platform, Apollo 3.
00:18:29.359 --> 00:18:35.200
CEO Jeff Cardinus told Reuters that real production versions were expected in 2027 and beyond.
00:18:35.440 --> 00:18:40.720
Aptronic has not published Apollo 3 specifications or a production volume target.
00:18:40.960 --> 00:18:47.599
1x start US Neo deliveries in 2026, with some buyers expected to wait longer.
00:18:47.839 --> 00:18:55.680
The company plans to expand to other markets in 2027, and its subscription plan is scheduled to follow early access sales.
00:18:55.920 --> 00:19:02.319
Its 2027 manufacturing target depends on the planned factory expansion and more automation.
00:19:02.640 --> 00:19:05.599
Final verdict Apollo for Enterprise Scale.
00:19:13.519 --> 00:19:23.359
Its wheel or leg choice, battery strategy, fleet software, Jable Partnership, and Worksite Data Program are aimed at the problems businesses face when they run a robot every day.
00:19:23.519 --> 00:19:26.400
But Apollo 2 is still a learning and pilot platform.
00:19:26.559 --> 00:19:30.319
Buyers should not mistake it for a fully proven commercial fleet product.
00:19:30.559 --> 00:19:35.519
For investors and consumer robotics watchers, NEO has the boulder mass market plan.
00:19:35.839 --> 00:19:43.279
It has a public price, a home first design, in-house manufacturing, and a published capacity target far above aptronix.
00:19:43.440 --> 00:19:45.839
But the execution risks are just as large.
00:19:46.079 --> 00:19:50.400
Customer delivery remains unconfirmed, household autonomy is still developing.
00:19:50.559 --> 00:19:55.599
Remote support creates privacy questions, and the factory target is not verified output.
00:19:55.839 --> 00:19:59.920
So the clearest answer is not that one robot wins every category.
00:20:00.240 --> 00:20:04.160
Apollo is the stronger design for scaling controlled enterprise work.
00:20:04.480 --> 00:20:08.799
NEO is the more ambitious design for bringing humanoids into homes at volume.
00:20:09.039 --> 00:20:21.440
The winner in the long run will be the company that turns those designs into reliable task completion, low service costs, clear safety evidence, and deliveries, not the company with the most eye-catching demonstration.
00:20:21.599 --> 00:20:25.359
All links to sources are available in the text version of this article.
00:20:25.519 --> 00:20:30.079
You can find the full article at robotcomparisons.com slash blog.
00:20:30.319 --> 00:20:31.359
Thanks for listening.
00:20:31.599 --> 00:20:33.680
Thanks for listening to Robot Comparisons.
00:20:33.759 --> 00:20:38.720
If you found this episode valuable, make sure to subscribe so you never miss a new deep dive.
00:20:38.880 --> 00:20:46.079
For the full written breakdown, detailed stuck sheets, and side by side comparisons, head over to robotcomparisons.com.
00:20:46.240 --> 00:20:52.000
That's R O B O T C O M P A R I S O N S.com.
00:20:52.240 --> 00:20:54.559
New research drops multiple times a week.
00:20:54.720 --> 00:20:56.240
See you in the next one.
00:00:00.000 --> 00:00:02.879
Aptronic Apollo vs One X NEO.
00:00:03.279 --> 00:00:06.480
Which 2026 humanoid has the more scalable design?
00:00:06.639 --> 00:00:09.759
By Akari Sato, Consumer Robotics Specialist.
00:00:09.919 --> 00:00:13.199
Status checked October 27, 2026.
00:00:13.439 --> 00:00:17.600
A humanoid robot can look ready for work and still be far from scalable.
00:00:17.839 --> 00:00:27.839
The real test is whether it can do useful jobs safely, reliably, at a cost buyers can defend, and whether the company can build, support, and improve a large fleet.
00:00:28.000 --> 00:00:32.320
Eptronix Apollo 2 and OneX's NEO are aimed at different markets.
00:00:32.560 --> 00:00:36.560
Apollo is being developed for factories, warehouses, and other businesses.
00:00:36.799 --> 00:00:40.320
NEO is designed for homes and is already listed for pre-order.
00:00:40.560 --> 00:00:49.439
That makes this comparison less like choosing between two versions of the same product and more like comparing two plans for getting humanoid robots into everyday use.
00:00:49.679 --> 00:00:54.399
The short verdict, Apollo 2, has the stronger design for scaling industrial work.
00:00:54.479 --> 00:01:01.520
Its wheeled and bipedal options, swappable batteries, fleet tools, and manufacturing partnership suit business operations.
00:01:01.759 --> 00:01:11.359
NEO has the clearer path to a mass market product, a published price, a lighter and softer body, and a company-run factory with ambitious capacity targets.
00:01:11.599 --> 00:01:18.000
But those targets are not the same as customer deliveries, and NEO's early service still includes remote human help.
00:01:18.159 --> 00:01:20.079
One distinction matters throughout.
00:01:20.319 --> 00:01:23.920
Apollo 2 is Apptronix's current learning and pilot platform.
00:01:24.159 --> 00:01:31.359
Aptronix says its next commercial platform will follow, with real production versions expected in 2027 and beyond.
00:01:31.599 --> 00:01:39.920
NEO, by contrast, is being offered for home pre-order, but as of October 27, OneX had not announced customer deliveries.
00:01:40.079 --> 00:01:45.040
At a glance for the full table, please open this article on robotcomparisons.com.
00:01:45.280 --> 00:01:50.799
The comparison is uneven because Aptronic has not published a full Apollo 2 specification sheet.
00:01:50.959 --> 00:02:00.400
NIO's numbers are easier to find, but many are company-listed specifications, not independent proof of work speed, safety, or reliability in customer homes.
00:02:00.640 --> 00:02:05.200
Two designs built around different work, Apollo puts choice of mobility first.
00:02:05.439 --> 00:02:08.639
Apollo 2 comes in wheeled and bipedal configurations.
00:02:08.800 --> 00:02:10.719
That is a practical design choice.
00:02:10.960 --> 00:02:15.919
On flat factory floors, wheels can avoid the extra balance and fall risk of walking.
00:02:16.159 --> 00:02:22.400
Legs are more useful where the robot must cross spaces built around people or handle routes that wheels cannot manage.
00:02:22.639 --> 00:02:28.879
Aptronix says the wheeled version is designed to fit with existing safety standards for industrial mobile robots.
00:02:29.120 --> 00:02:35.840
That is a design aim, not public proof that a particular Apollo 2 configuration has a safety certification.
00:02:36.000 --> 00:02:42.080
The bipedal version offers more flexibility, but also brings more demanding balance and safety questions.
00:02:42.400 --> 00:02:47.599
This modular approach makes Apollo's design adaptable to a business's site and task.
00:02:47.759 --> 00:02:55.360
It may also let Aptronic collect useful data on multiple body types without building a separate robot platform for every job.
00:02:55.520 --> 00:03:03.120
But buyers still need current dimensions, surface details, and performance data for the exact configuration they are considering.
00:03:03.360 --> 00:03:05.120
NEO puts home use first.
00:03:05.280 --> 00:03:06.960
NEO makes a different trade.
00:03:07.120 --> 00:03:10.639
It is a single bipedal form designed to live around people.
00:03:10.879 --> 00:03:22.000
Its 66-pound weight, soft outer body, washable suit, and quiet operation are sensible choices for a home, where people, pets, furniture, and clutter share the same space.
00:03:22.240 --> 00:03:26.560
Its design is more consumer-friendly than a typical industrial machine.
00:03:26.719 --> 00:03:30.080
But a home is not automatically an easy place for a robot.
00:03:30.240 --> 00:03:36.960
It has narrow paths, fragile objects, changing light, children, pets, and many tasks that are hard to describe in advance.
00:03:37.199 --> 00:03:44.800
NEO's lighter body may make it easier to live with, but does not by itself prove the robot can safely complete chores without supervision.
00:03:45.120 --> 00:03:47.439
Hardware, payload, and working time.
00:03:47.599 --> 00:03:48.560
Size and weight.
00:03:48.719 --> 00:03:54.800
NEO publishes the Clearer Case 1X lists NEO at 5 feet 6 inches and 66 pounds.
00:03:54.960 --> 00:04:04.560
Aptronic's older Apollo materials list a robot at about 5 feet 8 inches and 160 pounds, but Aptronic has not confirmed those figures for Apollo 2.
00:04:04.800 --> 00:04:09.919
Buyers should not treat the older body's weight or size as a current Apollo 2 specification.
00:04:10.159 --> 00:04:12.400
The difference matters for more than shipping.
00:04:12.639 --> 00:04:16.480
A heavy robot may need more careful handling during setup and service.
00:04:16.720 --> 00:04:21.759
A lighter robot is easier to move, but it may have different limits on strength and stability.
00:04:21.920 --> 00:04:26.800
Without current Apollo 2 measurements, buyers cannot make a fair apples to apples comparison.
00:04:27.040 --> 00:04:27.839
Actuators.
00:04:28.079 --> 00:04:29.839
Apollo emphasizes efficiency.
00:04:30.000 --> 00:04:32.079
Neo emphasizes gentle contact.
00:04:32.319 --> 00:04:34.639
Actuators are the parts that make a robot move.
00:04:34.879 --> 00:04:43.120
Aptronix says Apollo 2 uses its own patented actuators, designed for 90% plus energy efficiency, simple maintenance, and mass production.
00:04:43.279 --> 00:04:47.360
The company also says the design reduces reliance on single source suppliers.
00:04:47.600 --> 00:04:54.160
These are important scale goals, but they are company claims, not published fleet energy or repair results.
00:04:54.480 --> 00:04:56.480
NEO uses tendon drives.
00:04:56.639 --> 00:05:03.360
Motors pull tendons to move the joints, more like muscles pulling on limbs than a rigid motor sitting at every joint.
00:05:03.680 --> 00:05:12.560
One X says this design has low inertia and is backdrivable, meaning an external push can move a joint rather than meeting a hard, unyielding stop.
00:05:12.720 --> 00:05:16.079
Its product page lists 95% backdrivability.
00:05:16.319 --> 00:05:20.000
That approach fits a home robot where gentle contact matters.
00:05:20.240 --> 00:05:22.319
The strategic trade-off is clear.
00:05:22.560 --> 00:05:26.879
Apollo is framed around efficient industrial movement and maintainable hardware.
00:05:27.199 --> 00:05:31.279
NEO is framed around softer contact and close work with people.
00:05:31.439 --> 00:05:39.839
Neither has yet published the kind of long-term service data, such as real-world failure rates, that would settle which system is cheaper to keep running.
00:05:40.079 --> 00:05:40.720
Payload.
00:05:40.800 --> 00:05:44.160
Compare the useful load, not the biggest number NEO lists.
00:05:44.319 --> 00:05:50.000
154 pounds lift, 55 pounds carry, and 18 pounds arm payload.
00:05:50.240 --> 00:05:56.240
These figures do not mean that NEO can safely carry a 154-pound object around a home.
00:05:56.639 --> 00:06:01.360
For buyers, the arm payload and carry figures are more useful than the broad lift claim.
00:06:01.600 --> 00:06:10.399
Aptronic's older Apollo materials say the earlier robot could lift 55 pounds, but that is not a current Apollo 2 payload rating.
00:06:10.639 --> 00:06:13.680
Aptronic has not published one for the current version.
00:06:13.839 --> 00:06:22.639
Until it does, procurement teams should ask for limits by reach, speed, object shape, and whether the robot is walking or using a wheeled base.
00:06:23.040 --> 00:06:23.680
Runtime.
00:06:23.839 --> 00:06:25.759
Apollo has a clearer shift work plan.
00:06:25.920 --> 00:06:33.600
NEO lists a 4-hour runtime from an 842-watt-hour battery and says it can plug itself in when it needs charging.
00:06:33.839 --> 00:06:36.240
Its page also lists quick charging.
00:06:36.399 --> 00:06:43.120
These are useful specifications, but buyers still need to test how runtime changes with different tasks and loads.
00:06:43.360 --> 00:06:52.480
Aptronic describes swappable batteries, opportunity charging, and tethering as ways to keep Apollo working towards 7-day, 22-hour operation.
00:06:52.639 --> 00:06:59.360
That is an uptime design strategy, not a published guarantee that the robot will work reliably for 22 hours a day.
00:06:59.680 --> 00:07:04.240
Aptronic does not list a current Apollo 2 runtime or battery swap time.
00:07:04.480 --> 00:07:09.199
For a factory buyer, Apollo's battery plan is the more mature fit on paper.
00:07:09.360 --> 00:07:16.079
For Neo, self-charging is appealing in a home, but the robot still needs to navigate to its charger and dock successfully.
00:07:16.319 --> 00:07:20.399
Hands, sensing, and AI Neo has more published hand detail.
00:07:20.560 --> 00:07:24.319
The Neo products page lists 22 degrees of freedom per hand.
00:07:24.480 --> 00:07:34.480
In a July 2026 update, 1X described its new hand as having 25 degrees of freedom total, 22 in the fingers and palm, plus 3 at the wrist.
00:07:34.639 --> 00:07:42.720
That appears to be a difference in what is counted, not necessarily a conflict, but buyers should confirm which hand revision is included in the NEO they are offered.
00:07:42.959 --> 00:07:48.399
1X says the hands include force and touch sensing and are designed to bend under contact.
00:07:48.560 --> 00:07:57.839
That is promising for tasks like handling small or fragile objects, but a highly articulated hand only helps if the robot can use it reliably and quickly.
00:07:58.079 --> 00:08:01.439
1X has not published customer home success rates for chores.
00:08:01.839 --> 00:08:10.000
Apptronic calls Apollo 2's object handling dexterous, but does not publish a standard hand model or a full set of hand specifications for buyers.
00:08:10.240 --> 00:08:16.560
Google DeepMind has tested Apollo 2 with different hands, including the SharpaWave and Inspire hands.
00:08:16.720 --> 00:08:20.959
That shows Apollo can be used as a platform for AI and hardware experiments.
00:08:21.199 --> 00:08:25.920
It does not tell buyers which hand configuration will ship for a specific deployment.
00:08:26.160 --> 00:08:37.200
NEO publishes more of its sensors One X lists NIO with two stereo fisheye cameras, four microphones, Wi-Fi, Bluetooth, and 5G.
00:08:37.440 --> 00:08:42.240
It also lists an NVIDIA Jetson Thor computer inside its Neo Cortex system.
00:08:42.399 --> 00:08:48.080
This gives buyers a clearer view of the sensors and onboard compute in the advertised product.
00:08:48.320 --> 00:08:55.440
Apptronic's Apollo 2 materials describe perception systems, but do not provide a comparable current sensor list.
00:08:55.679 --> 00:09:04.480
That is an information gap for buyers who need to check camera coverage, sensor redundancy, data handling, or fit with site security rules.
00:09:04.799 --> 00:09:08.080
Both are learning systems, not finished all-purpose workers.
00:09:08.240 --> 00:09:15.759
Apptronic connects Apollo 2, its Artemis control system, Fleet Connect Fleet Software, and its Google DeepMind partnership.
00:09:15.919 --> 00:09:24.720
Robot Park, a nearly 90,000 square foot site, lets Apollo robots collect task data in logistics, manufacturing, and retail settings.
00:09:24.960 --> 00:09:28.639
Apptronic says that data helps train Gemini Robotics models.
00:09:28.960 --> 00:09:31.679
This is a strong learning and operations strategy.
00:09:31.919 --> 00:09:35.840
Robots work, data improves the models, and the fleet can be updated.
00:09:36.080 --> 00:09:41.600
Okay, Google's July 2026 Gemini Robotics 2 results show both promise and limits.
00:09:41.759 --> 00:09:44.960
In its tests, Apollo 2's success varied by task.
00:09:45.120 --> 00:09:54.559
For example, performance on whole body pickup tasks ranged from 45.7% for items on the floor to 76.3% for items on a shelf.
00:09:54.720 --> 00:09:58.000
On some multifinger hand tasks, results were uneven.
00:09:58.240 --> 00:10:06.639
Google reported 92% success for unscrewing a bulb, but 36% for screwing one in, and 40% for closing a zip bag.
00:10:06.879 --> 00:10:15.919
These are controlled tests, not factory shift results, but they are a useful reminder that a successful demonstration is not the same as reliable everyday performance.
00:10:16.240 --> 00:10:20.960
NEO's AI stack is described through Redwood AI and OneX's world model.
00:10:21.279 --> 00:10:26.080
The product is advertised as having basic autonomy and learning more skills over time.
00:10:26.240 --> 00:10:33.039
But NIO also offers scheduled expert mode, where a 1x worker remotely guides it through tasks it does not yet know.
00:10:33.200 --> 00:10:39.039
That makes NEO's early model a blend of autonomy and human help, not a fully independent housekeeper.
00:10:39.360 --> 00:10:43.759
Safety, privacy, and maintenance Apollo is built around controlled work zones.
00:10:44.000 --> 00:10:46.960
Apollo's safety plan includes hardware-level safety zones.
00:10:47.200 --> 00:10:58.879
Aptronix says its impact zone can pause movement when an object is detected within a defined radius, while a configurable perimeter zone changes how the robot responds to nearby people and objects.
00:10:59.039 --> 00:11:04.320
That is useful for structured workplaces where safety boundaries can be set and monitored.
00:11:04.480 --> 00:11:10.240
For buyers, the next step is to ask for the safety case for the exact task and configuration.
00:11:10.480 --> 00:11:16.720
What happens when a person enters the work area, a load shifts, or the robot loses balance or communication?
00:11:16.960 --> 00:11:22.639
The public product page does not provide independent test results or a full list of certifications.
00:11:22.960 --> 00:11:26.879
NEO is softer, but also brings sensors into private rooms.
00:11:27.039 --> 00:11:34.080
NEO's soft lattice body, covered joints, and low inertia tendon drives are meant to reduce injury risk during contact.
00:11:34.240 --> 00:11:39.200
1X also lists a head injury metric and says the body has no pinch points.
00:11:39.360 --> 00:11:45.360
These are design features, not proof that every home task is safe around children, pets, or vulnerable people.
00:11:45.600 --> 00:11:52.879
1X's own guidance says people should remain attentive and follow its safety instructions, especially around those groups.
00:11:53.200 --> 00:12:00.080
NEO's remote help raises a second safety question: who can see the home through the robot's cameras and when.
00:12:00.799 --> 00:12:03.919
1X presents expert support as scheduled.
00:12:04.080 --> 00:12:10.879
The company's FAQ, as reported in October 2026, says owners must accept a session.
00:12:11.120 --> 00:12:21.279
Even with that control, buyers should ask what sensor data is sent, whether sessions are recorded, how long data is kept, and how data is used to train models.
00:12:21.600 --> 00:12:25.279
Public information does not answer all of those questions clearly.
00:12:25.519 --> 00:12:27.759
Apollo brings a different data risk.
00:12:27.919 --> 00:12:36.480
Its data comes from workplaces and customer sites, where video or task records may expose production methods, layouts, or company information.
00:12:36.720 --> 00:12:43.200
Enterprise buyers should settle data ownership, access, retention, and model training rights in the contract.
00:12:43.360 --> 00:12:52.799
Maintenance is a scale test for both Apollo's swappable batteries and Apptronix claims of simpler, more maintainable actuators are built around reducing downtime.
00:12:53.039 --> 00:13:04.799
Jabil's role adds manufacturing and supply chain expertise, but public materials do not give buyers comparable service intervals, repair costs, or field failure rates for Apollo 2.
00:13:05.120 --> 00:13:12.159
NEO's washable outer suit and OneX's in-house manufacturing may simplify some repairs and replacement parts.
00:13:12.399 --> 00:13:17.600
The company also says its hands have been tested for millions of interaction cycles.
00:13:17.840 --> 00:13:27.039
That is a useful engineering signal, but buyers still need to know how often parts need service, who will do that work, and what it will cost after the warranty.
00:13:27.360 --> 00:13:28.879
How they fit the work.
00:13:29.120 --> 00:13:33.200
Factory and warehouse, Apollo is the stronger industrial choice today.
00:13:33.360 --> 00:13:36.799
Its wield base suits repeatable routes on level floors.
00:13:36.879 --> 00:13:41.679
Its bipedal version is the more flexible option for spaces designed around people.
00:13:41.919 --> 00:13:47.200
Swappable batteries and fleet management tools make the product direction fit business operations.
00:13:47.600 --> 00:13:52.799
Apollo says Apollo 2 is used at robot park and at partner or customer sites.
00:13:53.039 --> 00:13:58.639
Reuters reported that Apollo had built hundreds of Apollo 2 robots by June 2026.
00:13:59.039 --> 00:14:02.240
Well noting the company did not share how many were deployed.
00:14:02.480 --> 00:14:09.360
That is evidence of real platform activity, but it is not the same as proven production scale customer use.
00:14:09.679 --> 00:14:15.600
NEO is being built in a factory, but that does not yet show it can do factory work for a customer.
00:14:15.759 --> 00:14:22.639
1X says some NEOs are helping with internal factory tasks, including basic warehousing and logistics.
00:14:22.879 --> 00:14:25.200
Those activities are useful tests.
00:14:25.360 --> 00:14:30.159
They are not yet evidence of large, repeatable deployments at outside companies.
00:14:30.480 --> 00:14:38.480
Retail Apollo has the clearer business case for back-of-house retail tasks, such as moving stock or replenishing shelves.
00:14:38.720 --> 00:14:45.360
Aptronic lists retail as a target area and says robot park robots learn across retail tasks.
00:14:45.519 --> 00:14:53.440
But public sources do not show sustained retail results, such as cases handled per hour or how often a person must intervene.
00:14:53.759 --> 00:14:59.679
Neo's smaller, softer, quieter design may be more comfortable in a people-filled space.
00:14:59.840 --> 00:15:08.799
Yet its home first interface and early reliance on expert support make it hard to recommend for a retail operation without a dedicated pilot.
00:15:09.039 --> 00:15:14.080
Home Neo is the obvious fit by design, but not yet the proven home worker.
00:15:14.240 --> 00:15:17.759
It's quieter, lighter, and designed around home chores.
00:15:17.919 --> 00:15:24.320
Its voice interface, app, self-charging feature, and softbody all make sense for domestic use.
00:15:24.480 --> 00:15:35.039
However, OneX' product page describes basic autonomy for early owners, and the company had not announced customer deliveries as of October 27, 2026.
00:15:35.360 --> 00:15:40.559
Apollo could eventually serve homes, but it is not the consumer product in this comparison.
00:15:40.799 --> 00:15:49.440
Its current strengths and business model are directed at enterprise use, and there is no public home price or consumer ordering channel.
00:15:49.759 --> 00:15:53.519
Price, manufacturing, and the 2027 Outlook.
00:15:53.759 --> 00:15:55.600
NEO has a posted price.
00:15:55.919 --> 00:15:58.159
Apollo is an enterprise quote.
00:15:58.399 --> 00:16:06.159
One X lists NEO at$20,000 for early access ownership or$499 per month for a subscription.
00:16:06.320 --> 00:16:16.000
Its order page lists a$200 deposit and says US deliveries start in 2026, with early access customers first and subscriptions later.
00:16:16.240 --> 00:16:24.159
At$499 a month,$36 months would total$17,964 before any other costs.
00:16:24.399 --> 00:16:28.960
The terms and support package still matter more than that simple calculation.
00:16:29.360 --> 00:16:34.399
As of October 27th, NEO remained in the pre-order stage in public updates.
00:16:34.559 --> 00:16:44.399
1X said its early units were being used for internal development and home testing, and an October 2026 report found no announced customer delivery.
00:16:44.639 --> 00:16:53.600
The company still aims to begin deliveries in 2026, but a buyer should treat a specific arrival date as unconfirmed until 1X gives one.
00:16:53.919 --> 00:16:57.039
Aptronic does not publish an Apollo 2 price.
00:16:57.279 --> 00:17:00.720
Its site directs interested businesses to contact the company.
00:17:00.960 --> 00:17:05.599
Buyers should expect a custom enterprise process and ask for the full cost of a pilot.
00:17:05.839 --> 00:17:11.359
Robot, software, integration, service, training, safety work, and downtime.
00:17:11.680 --> 00:17:16.319
No public price estimate is dependable enough to use as a purchase benchmark.
00:17:16.559 --> 00:17:19.599
Manufacturing plans are not the same as output.
00:17:27.759 --> 00:17:34.240
Jable's global production network could help Atronics scale, but the companies have not announced a public Apollo production target.
00:17:34.559 --> 00:17:44.000
Aptronic also has a research partnership with Google DeepMind, while Mercedes Benz and GXO are among the customer or partner environments named for Apollo work.
00:17:44.400 --> 00:17:47.039
One X is taking a more vertically integrated route.
00:17:47.200 --> 00:18:02.240
It says its 58,000 square foot Hayward factory makes key parts in-house and has capacity for 10,000 NEOs per year, with a target of 100,000 plus units per year by the end of 2027 as automation and additional factory capacity grow.
00:18:02.480 --> 00:18:07.440
The company also says it sold out its first year of production capacity in five days.
00:18:07.599 --> 00:18:16.640
Those are major ambitions, but capacity and pre-order demand do not prove that 100,000 robots will be built, shipped, and supported in 2027.
00:18:16.960 --> 00:18:19.839
The near-term timeline is different for each company.
00:18:20.160 --> 00:18:21.039
Aptronic.
00:18:21.200 --> 00:18:23.440
Continue pilots through 2026.
00:18:23.680 --> 00:18:29.119
The company says Apollo 2 data is feeding the next commercial platform, Apollo 3.
00:18:29.359 --> 00:18:35.200
CEO Jeff Cardinus told Reuters that real production versions were expected in 2027 and beyond.
00:18:35.440 --> 00:18:40.720
Aptronic has not published Apollo 3 specifications or a production volume target.
00:18:40.960 --> 00:18:47.599
1x start US Neo deliveries in 2026, with some buyers expected to wait longer.
00:18:47.839 --> 00:18:55.680
The company plans to expand to other markets in 2027, and its subscription plan is scheduled to follow early access sales.
00:18:55.920 --> 00:19:02.319
Its 2027 manufacturing target depends on the planned factory expansion and more automation.
00:19:02.640 --> 00:19:05.599
Final verdict Apollo for Enterprise Scale.
00:19:13.519 --> 00:19:23.359
Its wheel or leg choice, battery strategy, fleet software, Jable Partnership, and Worksite Data Program are aimed at the problems businesses face when they run a robot every day.
00:19:23.519 --> 00:19:26.400
But Apollo 2 is still a learning and pilot platform.
00:19:26.559 --> 00:19:30.319
Buyers should not mistake it for a fully proven commercial fleet product.
00:19:30.559 --> 00:19:35.519
For investors and consumer robotics watchers, NEO has the boulder mass market plan.
00:19:35.839 --> 00:19:43.279
It has a public price, a home first design, in-house manufacturing, and a published capacity target far above aptronix.
00:19:43.440 --> 00:19:45.839
But the execution risks are just as large.
00:19:46.079 --> 00:19:50.400
Customer delivery remains unconfirmed, household autonomy is still developing.
00:19:50.559 --> 00:19:55.599
Remote support creates privacy questions, and the factory target is not verified output.
00:19:55.839 --> 00:19:59.920
So the clearest answer is not that one robot wins every category.
00:20:00.240 --> 00:20:04.160
Apollo is the stronger design for scaling controlled enterprise work.
00:20:04.480 --> 00:20:08.799
NEO is the more ambitious design for bringing humanoids into homes at volume.
00:20:09.039 --> 00:20:21.440
The winner in the long run will be the company that turns those designs into reliable task completion, low service costs, clear safety evidence, and deliveries, not the company with the most eye-catching demonstration.
00:20:21.599 --> 00:20:25.359
All links to sources are available in the text version of this article.
00:20:25.519 --> 00:20:30.079
You can find the full article at robotcomparisons.com slash blog.
00:20:30.319 --> 00:20:31.359
Thanks for listening.
00:20:31.599 --> 00:20:33.680
Thanks for listening to Robot Comparisons.
00:20:33.759 --> 00:20:38.720
If you found this episode valuable, make sure to subscribe so you never miss a new deep dive.
00:20:38.880 --> 00:20:46.079
For the full written breakdown, detailed stuck sheets, and side by side comparisons, head over to robotcomparisons.com.
00:20:46.240 --> 00:20:52.000
That's R O B O T C O M P A R I S O N S.com.
00:20:52.240 --> 00:20:54.559
New research drops multiple times a week.
00:20:54.720 --> 00:20:56.240
See you in the next one.
