When evaluating which company is furthest along in deploying humanoid robots on live factory and warehouse floors, Agility Robotics holds the longest track record of continuous commercial operation , though Figure AI leads in high-volume automotive production integration, and Tesla is scaling internal hardware manufacturing the fastest.
Longest commercial tenure: Nearly 3 years of continuous multi-facility deployment; Digit 4 logged over 65,000 operational hours. Mass-manufacturing purpose-built at "RoboFab".
Figure AI
Figure 02 / Figure 03
BMW Manufacturing (Spartanburg, SC)
Deepest production line integration: Deployed on a live automotive floor, contributing to the assembly of over 30,000 vehicles handling sheet metal and complex bin-sequencing.
Tesla
Optimus (Gen 2 / Gen 3)
Tesla Fremont & Giga Texas
Largest internal pilot & scale: Hundreds to ~1,000 units deployed internally for battery sorting and kitting/data collection, with massive component scaling underway.
Boston Dynamics
All-New Electric Atlas
Hyundai Motor Group
Ramping production: Production underway at headquarters; 2026 output committed to Hyundai metaplants and AI testing with Google DeepMind ahead of full commercial rollout.
Breakdown of the Leaders
Agility Robotics (Best for cumulative operational hours):
Digit has logged over 65,000 hours of actual work across commercial customer facilities like GXO, Amazon, and Toyota. They are the furthest along in generalized logistics and repetitive material handling, and their dedicated manufacturing plant ("RoboFab") is scaled to produce thousands of units.
Figure AI (Best for live automotive manufacturing):
Figure has proven its capability directly on automotive production lines. Their robots spent extended runs inside BMW’s Spartanburg plant assisting in live manufacturing tasks—moving past simple demonstrations to sequence unsorted parts and handle sheet metal for tens of thousands of consumer vehicles.
Tesla (Best for internal volume and manufacturing ramp):
While Tesla has mostly utilized Optimus internally for data collection and factory kitting tasks (such as sorting 4680 battery cells) rather than full commercial labor substitution, its hardware production pipeline is ramping aggressively toward thousands of units per quarter.
If you'd like to narrow this down, let me know:
Are you more interested in logistics/warehousing or precision assembly lines?
Do you want a deeper look at the AI software (VLA models) driving these robots?
As of September 2026, the strongest evidence points to Agility Robotics as the company furthest along in actual sustained commercial deployment of humanoids in industrial environments.
The important nuance is that “factory floors” specifically produces a slightly different answer.
Agility Robotics — Digit: Agility says Digit has accumulated 65,000+ operating hours, nearly three years of continuous commercial deployment, across customers including GXO, Schaeffler, Amazon, and Toyota Motor Manufacturing Canada. Toyota signed a commercial Robots-as-a-Service agreement in February 2026 following a pilot.
Figure AI — BMW: Figure has perhaps the strongest automotive factory deployment. BMW says Figure 02 worked at its Spartanburg plant throughout 2025, contributing to production of 30,000+ BMW X3s, moving 90,000+ components over about 1,250 operating hours. Figure 03 returned to the plant in June 2026 for a substantially more complicated logistics/sequencing task.
Apptronik — Apollo: Apollo is deployed in manufacturing-related pilots with Mercedes-Benz and Jabil and has expanded into customer sites, but the publicly documented deployments are not yet at Agility's duration/scale.
So, if I had to define the race by deployment maturity
One particularly telling data point: Agility announced in September 2026 that Digit had been operating commercially for nearly three years, while reporting more than 65,000 hours across multiple industrial sites.
Bottom line: If you mean “Who has put humanoids into real industrial operations at scale and for the longest time?” → Agility Robotics/Digit. If you mean “Who has demonstrated the most compelling humanoid use case directly on an automotive production line?” → Figure AI/BMW is arguably the more relevant benchmark, based on the publicly documented deployments.
If you're evaluating this from an investment/industry standpoint, I can also compare Agility vs Figure vs Apptronik vs Tesla on units deployed, customer contracts, manufacturing capacity, autonomy, and likely path to mass production.
When evaluating which company is furthest along in deploying humanoid robots on actual commercial and factory floors, Agility Robotics and Figure lead in verified operational hours and live multi-facility deployments, while Tesla pursues the most aggressive vertical integration and internal manufacturing scale.
The industry breaks down into distinct approaches:
Agility Robotics (The Commercial Leader in Logistical Hours): Agility is arguably furthest along in cumulative real-world commercial operations. Their previous-generation Digit 4 humanoid logged over 65,000 hours of actual work in live customer environments—such as handling totes at GXO Logistics , and testing at Amazon and Toyota . They have since unveiled Digit 5 , backed by hundreds of millions in multi-year customer orders, designed to work safely alongside humans without safety cages.
Figure (The Leader in Automotive Plant Integration): Figure has moved rapidly through high-profile, rigorous manufacturing pilots. Following a successful pilot with its Figure 02 model positioning sheet metal parts at BMW’s Spartanburg, South Carolina plant, Figure upgraded to the Figure 03 model to take on complex sequencing tasks on BMW’s active factory floor. They have also secured commercial deployment deals outside automotive (such as Catalyst Brands), proving rapid transition from pilot to tougher logistical duties.
Tesla (The Scale & Internal Manufacturing Ambition): Tesla has transitioned its Fremont, California facility toward internal manufacturing lines for Optimus and utilizes early units internally. However, while Tesla commands massive long-term ambitions (such as retooling factories for extreme high-volume output and constructing dedicated lines at Giga Texas), verified external, multi-customer commercial deployments on third-party factory floors remain in early ramp-up phases compared to the accrued commercial runtime demonstrated by Agility and Figure.
If you'd like to explore this further, let me know if you want to look at:
A deeper breakdown of hardware specs (payloads and runtimes)
The specific use cases (welding vs. tote-moving/sequencing)
As of September 2026, the strongest evidence points to Agility Robotics as the company furthest along in actual sustained industrial deployment—particularly if you mean robots doing productive work on factory/warehouse floors rather than demonstrations or short pilots.
Agility / Digit: Agility says Digit has accumulated 65,000+ hours of operation, with nearly three years of continuous commercial deployment across customers including Toyota, Schaeffler, Amazon and GXO. Toyota Motor Manufacturing Canada has moved from a pilot to a commercial Robots-as-a-Service agreement.
Figure AI: Probably the closest competitor for automotive factory work. BMW says Figure 02 operated for about 11 months at its Spartanburg plant, working 10-hour shifts and moving 90,000+ components during roughly 1,250 operating hours. BMW subsequently began deploying the newer Figure 03 for logistics work.
Apptronik / Apollo: Has commercial agreements with Mercedes-Benz and GXO, but the publicly documented deployments are less mature than Agility's accumulated operating history.
Tesla Optimus: Despite the enormous attention, I would not put Tesla first based on publicly documented external factory deployment. The distinction is between robots operating in Tesla's own facilities and commercially deployed fleets at independent customers.
The important distinction
If by "factory floors" you specifically mean automotive manufacturing rather than warehouses/logistics, then Figure AI currently has the strongest documented case: BMW has already run Figure robots through sustained production work and is moving to a second-generation deployment.
So my factual snapshot would be:
Company
Where it stands
Agility Robotics
Most mature sustained commercial industrial deployment
Figure AI
Most compelling automotive factory deployment
Apptronik
Commercial factory programs, scaling phase
Tesla
Significant internal development/deployment, but less externally documented
One caveat: the International Federation of Robotics says humanoids are still a very small fraction of industrial robotics, and many units sold in 2025 were for R&D rather than productive labor. So "furthest along" here means relative maturity of real-world deployment, not that humanoids are already broadly replacing conventional factory robots.
If you're evaluating this from an investment/industry standpoint, the more interesting question is probably which company has the clearest path from dozens of pilots to thousands of economically useful robots. That produces a somewhat different comparison.
If by “furthest along” you mean actual humanoid robots doing productive work on real factory floors—not demos or pilot announcements—the clearest answer as of September 2026 is Agility Robotics with Digit.
Agility Robotics / Digit: Agility says Digit has accumulated 65,000+ operating hours, has been in commercial industrial deployment for nearly three years, and is working across customer sites including GXO, Schaeffler, Amazon, and Toyota Motor Manufacturing Canada. At GXO, Digit has handled 100,000 totes with about 98% on-task accuracy. Toyota has also moved from a pilot to a commercial Robots-as-a-Service agreement.
Figure AI / BMW: Figure has perhaps the most notable automotive manufacturing deployment. BMW says Figure 02 operated at its Spartanburg plant for ten months, supporting production of 30,000+ BMW X3s and working ten-hour shifts, five days a week.
Apptronik / Apollo: Apollo has substantial real-world activity, including manufacturing-related work with Mercedes-Benz and GXO, but the publicly documented deployment footprint and accumulated operating time currently trail Digit's.
The important distinction
Company
Humanoid
Factory/industrial deployment
Evidence of scale
Agility Robotics
Digit
Commercial deployments
65,000+ hours; multiple customers; commercial Toyota agreement
Figure AI
Figure 02
BMW production
So, Agility looks furthest along overall in sustained commercial deployment, while Figure has an especially strong case for depth of deployment inside an automotive factory.
One caveat: these companies disclose different metrics, so there isn't a standardized industry scoreboard. “65,000 hours” versus “30,000 vehicles” isn't an apples-to-apples comparison.
When looking at which company is furthest along in terms of actual, continuous commercial deployment and logged operational hours on real manufacturing and warehouse floors, Agility Robotics holds the longest track record. However, Tesla and Figure AI are neck-and-neck or leading in terms of raw scale and high-speed automotive factory integration.
The front-runners approach this milestone differently:
Agility Robotics (Digit): Furthest along in tenure and cumulative operational hours . Its bipedal robot, Digit, has logged over 65,000 hours of real-world operational work across commercial customer sites (including GXO, Amazon, Schaeffler, and Toyota) spanning nearly three years. With the rollout of their next-gen Digit 5 , they remain the benchmark for logged multi-customer industrial deployment data.
Tesla (Optimus): Furthest along in internal volume and high-speed manufacturing integration . Tesla has deployed large numbers of its Optimus units directly inside its own facilities (such as Gigafactory Texas and Fremont) to handle sorting, component transport, and factory-floor tasks.
Figure AI (Figure 03 / BMW Partnership): Furthest along in collaborative automotive manufacturing pilots . Figure’s ongoing partnership with the BMW Group at Plant Spartanburg in South Carolina has transitioned from initial proofs-of-concept to complex sequencing tasks—where humanoids pick, sort, and sequence irregular parts directly alongside human assembly teams.
If you'd like, I can break down:
The specific metrics (payload capacity, speed, and uptime) of Digit 5 vs. Optimus Gen 3
The economic return on investment (ROI) for deploying humanoids versus traditional stationary industrial arms
Let me know which angle or company you want to explore further.
When assessing which company is furthest along in deploying humanoid robots on actual factory floors (rather than just showing controlled lab demos), Figure AI and Tesla stand out as the frontrunners, though they approach deployment with different strategies and transparency.
A breakdown of where the key players stand reveals distinct differences in their real-world progress:
Figure AI (The Leader in Commercial/External Validation)
Current Status: Figure is widely considered the furthest along in commercial, audited industrial deployment with external manufacturing partners.
Factory Milestone: Following a successful pilot with the Figure 02 robot at the BMW Group Plant Spartanburg (which logged over 1,250 operating hours, moved 90,000+ components, and helped assemble 30,000 vehicles), Figure deployed its upgraded Figure 03 robot to the plant.
The Task: Moving beyond simple pick-and-place tasks, Figure 03 handles complex "sequencing" logistics—sorting unsorted components into trolleys for just-in-time delivery to the active assembly line. They have also established benchmark unit economics (such as a publicized $25/operating-hour metric).
Tesla (The Leader in Internal Scale & Ambition)
Current Status: Tesla utilizes its own factories (Fremont and Giga Texas) as the testing and deployment ground for the Optimus robot.
Factory Milestone: Tesla has deployed hundreds of internal units performing rudimentary factory tasks—such as sorting battery cells, handling parts, and doing basic kitting between stations.
The Catch: While Tesla boasts massive internal numbers and manufacturing scale, a large portion of their deployment is treated as an internal "Optimus Academy" for data collection and AI training rather than external commercial manufacturing contracts.
Other Notable Industrial Contenders
Boston Dynamics: Partnered with Hyundai to test the electric Atlas on automotive production lines, featuring autonomous battery-swapping capabilities.
Apptronik: Has active industrial pilot programs for its Apollo robot with companies like Mercedes-Benz and Jabil.
Unitree & AgiBot: Dominating global shipment volume numbers (particularly driven by mass production in China), though many of these units are sold to research institutions, developers, and logistics testbeds rather than heavy commercial automotive deployment.
Summary Verdict
If measured by proven, multi-month commercial operation inside a tier-one automaker's active production line, Figure AI holds the public lead. If measured by raw number of units deployed inside a proprietary manufacturing ecosystem, Tesla is operating at a comparable or larger scale, albeit mostly behind closed doors for internal R&D.
Would you like to dive deeper into the unit economics/costs per operating hour or compare the technical AI architectures (neural end-to-end vs. teleoperation hybrids) of Figure and Tesla?
As of September 2026, I’d put Agility Robotics furthest along in actual commercial deployment of humanoids on industrial floors, with its Digit robot.
The important caveat is that “furthest along” depends on what you measure:
🥇 Agility Robotics — Digit: Probably the leader in real-world commercial operating hours and number of customer sites. Digit entered commercial operations at GXO in June 2024, and Agility now says it is deployed in manufacturing, distribution and logistics. Its current customers include GXO, Amazon, Toyota Motor Manufacturing Canada, Schaeffler and Mercado Libre, with more than 65,000 hours of real-world operation reported.
🥈 Figure AI — Figure 03: Arguably the strongest case for actual automotive production. Figure's robots have been used at BMW Spartanburg, including production-line sheet-metal handling. BMW has now demonstrated Figure 03 in another production use case at Spartanburg.
🥉 Apptronik — Apollo: Very serious contender in automotive manufacturing. Mercedes-Benz has been piloting Apollo in its factories, with applications including moving components and supplying production stations. The commercial agreement dates back to 2024.
Tesla — Optimus: Potentially the biggest long-term threat, because Tesla can manufacture at enormous scale, but it isn't yet the leader in externally demonstrated factory deployment.
Boston Dynamics — Atlas: Technologically impressive, but definitely not the leader in deployment yet. Reuters reported this month that Atlas has not been deployed at scale.
My ranking
Company
Robot
Factory maturity
Agility Robotics
Digit
⭐⭐⭐⭐⭐
Figure AI
Figure 03
⭐⭐⭐⭐½
Apptronik
Apollo
⭐⭐⭐⭐
If your question is “Who has the most humanoid robots actually doing productive work today?” → Agility.
If it's “Who has demonstrated the most compelling humanoid use inside a true automotive production line?” → Figure is arguably ahead.
And there's an important distinction: none of these companies has yet demonstrated a general-purpose humanoid replacing a human worker across a broad range of factory tasks. Today's deployments are still highly constrained tasks—moving totes, handling parts, feeding stations, etc. Reuters' recent reporting on the Chinese market makes the same broader point: the industry is still struggling with dexterity, autonomy and adapting to unpredictable factory environments.
If you're asking this from an investment perspective, I'd rank the companies differently—Agility, Figure, Apptronik and Tesla have very different risk/reward profiles.
UBTECH and Agility Robotics are furthest along in commercial factory and warehouse deployments , leading a wave of real-world integration rather than just lab testing.
While Western media focuses heavily on conceptual pilots, several key players have crossed the threshold into active, multi-unit commercial deployments:
1. UBTECH (Walker S Series) — Most Units Deployed in Manufacturing
The Scale: UBTECH has delivered over 1,000 units of its Walker S/S2 humanoid robots, backed by massive commercial orders exceeding $195 million.
Factory Floor Presence: Extensively deployed across major Chinese automotive and electronics manufacturing lines including BYD, Geely Auto, FAW-Volkswagen, Dongfeng, and Foxconn.
The Work: Performing sorting, transport logistics, quality inspection, and assembly assistance side-by-side with human workers.
2. Agility Robotics (Digit) — Most Logistical Operating Hours
The Scale: Agility has secured over $300 million in multi-year orders for its bipedal Digit robots and scaled live deployments with logistics giants and manufacturers like Amazon and GXO.
Factory Floor Presence: Digit has logged extensive operational milestones, such as moving over 100,000 totes in live commercial warehouse and logistics routing environments.
The Work: Specializing heavily in material handling, moving totes, and repetitive trailer-loading/unloading tasks.
3. Figure AI (Figure 02 / Figure 03) — Fastest Integration in Western Automotive
The Scale: Partnered deeply with BMW at the Spartanburg manufacturing plant.
Factory Floor Presence: Following successful trials of the Figure 02, advanced variants like the Figure 03 are actively sorting complex vehicle components into sequencing trolleys directly on automotive production lines.
4. Other Notable Deployments
AgiBot: Highly prolific in total production volume in China (producing thousands of units spanning mixed industrial and commercial use).
Apptronik (Apollo): Actively deployed testing and moving assembly kits across Mercedes-Benz production facilities.
Tesla (Optimus): Rapidly scaling internal deployment across Tesla's own manufacturing plants, though lagging behind UBTECH and Agility in third-party commercial delivery scale.
Would you like to explore the unit economics and hourly operating costs of these deployed humanoid robots compared to human labor?
If you mean actual, sustained work on factory floors—not demos or announced production targets—Agility Robotics and Figure AI are currently the leaders, with Agility probably furthest ahead in commercial deployment overall.
My ranking
Agility Robotics — Digit
Digit is already deployed commercially at Amazon, GXO, Schaeffler, Mercado Libre, and Toyota Motor Manufacturing Canada.
It is doing relatively narrow but economically useful logistics/manufacturing jobs such as moving totes and bins.
Agility says its humanoids are deployed today in manufacturing, distribution and logistics.
Recent reporting says Digit has moved 100,000+ totes at GXO, and Toyota has signed a commercial Robots-as-a-Service agreement.
Figure AI — Figure 02/03
Figure has perhaps the most impressive demonstrated automotive-manufacturing deployment.
Its BMW Spartanburg trial ran for roughly 11 months; Figure 02 handled sheet-metal parts for welding and reportedly contributed to production of 30,000+ vehicles.
So if your definition of "furthest along" is complex manipulation directly inside an automotive production process, I'd give Figure the edge.
Apptronik — Apollo
Apollo has moved beyond demonstrations into Mercedes-Benz manufacturing operations, with the companies pursuing broader production-line use.
It has strong industrial backing and a major manufacturing focus, but its publicly demonstrated deployment footprint is still behind Agility/Figure.
Tesla — Optimus
Potentially the biggest long-term threat because Tesla can manufacture at enormous scale and use its own factories as a proving ground.
But today I wouldn't call Tesla the leader in deployed factory humanoids. Current evidence points to Figure and Agility having substantially stronger verified real-world deployment records.
The industry is still much earlier than headlines imply. Even the leaders are generally doing specific repetitive workflows, rather than replacing a human worker across an entire factory. Reuters recently reported that many Chinese humanoids, despite impressive hardware demonstrations, still aren't ready for genuinely autonomous factory work.
If I had to pick one company to watch for who will actually have thousands of humanoids doing productive factory work first, I'd pick Agility today—but Figure may have the stronger shot at being the first to crack genuinely complex manufacturing tasks.