Skip to main content

Heavy-Payload Industrial Robots Compared: Payload, Reach, Wrist Ratings

Heavy-payload industrial robots from 500 kg to 2,300 kg can be compared on their published payload, reach, wrist moment, wrist inertia, and mass, but only after each figure is read with the label and conditions its manufacturer printed. UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for aerospace and heavy industry from its Spokane Valley, WA facility, integrating Allen-Bradley PLC and motion control with in-house CNC machining, heat treating, and stress relief. This article sets the published specifications of heavy-payload models from FANUC, ABB, and Yaskawa side by side, explains where the columns do not line up, and shows how a buyer builds a shortlist for a specific heavy load. The robot sits at the drives and controls links of the chain, design → engineering → parts machining → fabrication → assembly → weld fatigue → stress relief → drives → controls → tuning → monitoring, and the model chosen fixes the base, the tooling budget, and the controls the rest of the cell is built around.

What separates the payload tiers of heavy industrial robots?​

The models compared here can be grouped into three rough tiers by maximum wrist load:

  • About 500 kg. FANUC's M-950iA/500 carries 500 kg at 2,830 mm reach. ABB's IRB 8700-550/4.20 carries 550 kg at 4.20 m.
  • About 800 to 1,000 kg. ABB's IRB 8700-800/3.50 carries 800 kg at 3.50 m. FANUC's M-1000iA carries 1,000 kg at 3,253 mm. Yaskawa's MH900 carries 900 kg at 4,683 mm horizontal reach.
  • Above 1,000 kg. FANUC's M-2000 family has four models rated 900 kg to 2,300 kg, three of them above 1,000 kg, at 3,734 mm or 4,683 mm reach depending on the model.

The tiers are a reading aid, not a standard classification. Among the models compared here, a step up in payload can mean a different arm, a different base, and a different wrist rating rather than a stronger version of the same robot: a buyer who needs 1,100 kg is past every model rated 1,000 kg or less, and that load moves into the M-2000 tier, where the lightest published model weighs 8,600 kg without its controller (FANUC RM-950iA catalog 2023, p. 2; ABB 3HAC087213-001 Rev. H 2026, §1.1.2 p. 12; FANUC RM-1000iA catalog 2022, p. 2; Yaskawa America DS-924 2021, p. 2; FANUC RM-2000 catalog 2024, p. 2).

How do heavy-payload models compare on payload, reach, and mass?​

The table sets out the figures each manufacturer prints. Reach and mass are given as printed, and the notes column records what each source says about its figures or construction.

ModelMax wrist loadReachRobot massSource note
FANUC M-950iA/500500 kg2,830 mm2,500 kgMass without controller
ABB IRB 8700-550/4.20550 kg4.20 m4,750 kgReach at wrist centre; mass excludes dress pack
ABB IRB 8700-800/3.50800 kg3.50 m4,750 kgReach at wrist centre; mass excludes dress pack
Yaskawa MH900900 kg4,683 mm horizontal, 6,209 mm vertical10,000 kgParallel-link construction
FANUC M-1000iA1,000 kg3,253 mm5,300 kgMass without controller
FANUC M-2000/900F-47A900 kg4,683 mm9,600 kgMass without controller
FANUC M-2000/1200F-37A1,200 kg, 1,350 kg option3,734 mm8,600 kgMass without controller
FANUC M-2000/1700F-47A1,700 kg4,683 mm12,500 kgMass without controller
FANUC M-2000/2300F-37A2,300 kg3,734 mm11,000 kgMass without controller

Two cautions apply. First, ABB states its reach at the wrist centre, and the reference point behind each other sheet's reach figure is read from that manufacturer's working-range drawing before the figures are compared or a layout is checked. Second, an older FANUC catalog, dated 2022, gives the M-2000iA/1200, /2300, /900L, and /1700L the same specification-table figures that the 2024 catalog gives the M-2000/1200F-37A, /2300F-37A, /900F-47A, and /1700F-47A, so a quote and a catalog may use different names for arms with identical published figures (FANUC RM-2000 catalog 2024, p. 2; FANUC RM-1000iA catalog 2022, p. 2; FANUC RM-950iA catalog 2023, p. 2; ABB 3HAC087213-001 Rev. H 2026, §1.1.2 p. 12, §1.1.3 p. 13, and §1.4 p. 23; Yaskawa America DS-924 2021, p. 2).

How do wrist moment and inertia ratings compare across models?​

Wrist ratings can decide a heavy-handling selection, because a long gripper or offset part can reach the wrist limits before the payload limit. FANUC and Yaskawa both print an allowable moment and an allowable inertia for each of the last three axes:

ModelAllowable wrist moment, axes 4 / 5 / 6Allowable wrist inertia, axes 4 / 5 / 6
FANUC M-950iA/5003,420 / 3,420 / 1,850 N·m1,100 / 1,100 / 445 kg·m²
FANUC M-1000iA8,800 / 8,800 / 5,800 N·m1,750 / 1,750 / 840 kg·m²
Yaskawa MH900, R / B / T14,700 / 14,700 / 4,900 N·m3,000 / 3,000 / 2,200 kg·m²
FANUC M-2000/1200F and 900F14,700 / 14,700 / 4,900 N·m2,989 / 2,989 / 2,195 kg·m²
FANUC M-2000/2300F and 1700F29,400 / 29,400 / 8,820 N·m7,500 / 7,500 / 5,500 kg·m²

Reading the table needs three notes:

  • Axis names differ. Yaskawa's sheet labels its six axes S, L, U, R, B, and T, with R as arm roll, B as wrist bend, and T as tool flange; R, B, and T are the fourth to sixth axes, the positions J4, J5, and J6 hold on a FANUC sheet.
  • One catalog's inertia conversions do not match. The M-950iA/500 catalog also gives inertia in kgf·cm·s², and those figures do not match its SI values, so only the kg·m² figures are used here.
  • Ratios tell more than totals. The M-1000iA carries twice the M-950iA/500's payload with about 2.6 times its J4 and J5 moment rating, while its J6 inertia rises from 445 to 840 kg·m², under twice. A load that spins about the flange axis can reach the J6 limit first.

The M-2000/1200F and the MH900 print the same moment figures, but matching figures on two sheets do not make the robots interchangeable; their inertia ratings, reach, and mass differ (FANUC RM-950iA catalog 2023, p. 2; FANUC RM-1000iA catalog 2022, p. 2; FANUC RM-2000 catalog 2024, p. 2; Yaskawa America DS-924 2021, p. 2).

Why do load-diagram ratings not fit the same wrist-rating columns?​

ABB presents its IRB 8700 wrist limits differently, and forcing them into the FANUC and Yaskawa columns would misstate them. ABB gives load diagrams, which plot allowable load against CG position; two wrist-inertia formulas, with limits of 1,100 kg·m² on axis 5 and 725 kg·m² on axis 6; and maximum permissible wrist torques due to payload that it states are for reference only and should not be used to calculate the permitted CG offset. The formulas, the torque figures, and a worked load case are set out in Sizing an Industrial Robot: Payload, Reach, Wrist Moment, and Inertia. The load diagrams assume a load inertia of 100 kg·m² and 50 kg on the upper arm housing.

So an IRB 8700 row in a moment column would show a torque the manufacturer says should not be used to calculate the permitted CG offset, and an inertia column would show a limit that is really the right-hand side of a formula. The comparison that holds is load case against load case: take the actual gripper and part, run the ABB formulas and load diagram for the ABB candidate, and compare the result with the other candidates' printed ratings for the same load (ABB 3HAC087213-001 Rev. H 2026, §1.6.1 p. 38, §1.6.3 pp. 45-46, and §1.6.4 p. 47).

Why can repeatability figures not be compared one-for-one?​

The accuracy figures on the sheets compared here carry different labels, and not every sheet states a test standard or test conditions, so a smaller number is not automatically a better robot:

ModelPrinted figureLabel and basis as printed
ABB IRB 8700-800/3.500.05 mmISO 9283 pose repeatability, at rated maximum load, maximum offset, 1.6 m/s; average of measurements on a small number of robots
ABB IRB 8700-550/4.200.08 mmSame basis
FANUC M-950iA/500±0.08 mm"Positioning accuracy," compliant with ISO 9283
FANUC M-1000iA±0.10 mm"Positioning accuracy," compliant with ISO 9283
FANUC M-2000/1200F and 900F±0.18 mm"Repeatability," compliant with ISO 9283
FANUC M-2000/2300F and 1700F±0.27 mm"Repeatability," compliant with ISO 9283
Yaskawa MH900±0.5 mm"Repeatability," no test standard stated on the sheet

ISO 9283 is the international standard for robot performance criteria and test methods, and ISO 9946 specifies how a manufacturer presents a robot's characteristics. A 1999 handbook chapter on industrial robotics standards by a NIST author states that the ISO 9283 tests are primarily intended for developing and verifying an individual robot's specifications, for prototype testing, or for acceptance testing, that the ISO subcommittee did not intend the standard for comparing robots of similar capacity and size, that its annex of standard test path lengths, loads, and velocities is optional, and that the US standard of that time, ANSI/RIA R15.05-1-1990, and ISO 9283 computed positioning repeatability with different formulas.

The practical rules are to quote each figure with its printed label, never to set a "positioning accuracy" value against a "repeatability" value, and, where placement matters, to write the acceptance test into the purchase specification (ISO 9283:1998; ISO 9946:1999; Dagalakis 1999, pp. 447-459, DOI 10.1002/9780470172506.ch24; ABB 3HAC087213-001 Rev. H 2026, §1.9.2 p. 59; FANUC RM-2000 catalog 2024, p. 2; Yaskawa America DS-924 2021, p. 2).

How do wrist-down and arm-mounted load allowances change the comparison?​

Two allowances sit outside the headline payload and can move a model up or down a shortlist.

The first is a wrist-down rating. ABB's IRB 8700 specification gives a separate "vertical wrist" load diagram for tasks where the wrist points down within ±10° of vertical. For the wrist pointing straight down, at 0° from vertical, ABB lists a maximum of 1,000 kg for the -800/3.50, with Zmax 0.662 m and Lmax 0.297 m, and 700 kg for the -550/4.20, with Zmax 0.602 m and Lmax 0.196 m. The -800/3.50 figure is 1,000 kg in Revision H of the OmniCore specification; the revision log of ABB's IRC5-controller product specification for the IRB 8700 records, at Revision A, that the value was changed from 950 kg to 1,000 kg, so a copy of that edition from before Revision A shows the lower figure. A top-down pick-and-place task may therefore fit an 800 kg-rated arm with a 1,000 kg load, while a task that tilts the wrist more than 10° from vertical is outside the vertical-wrist diagram.

The second is the allowance for equipment mounted on the arm. FANUC's M-950iA/500, M-1000iA, and M-2000 catalogs each allow 550 kg on the J2 base and 50 kg on the J3 arm or casing. ABB's load diagrams assume 50 kg on the upper arm housing. Valve banks, vacuum generators, and dress packs are placed and counted against these allowances, not simply added to the wrist payload.

The failure mode is quoting a wrist-down figure for a task that also rotates the part to a side-facing pose, where the vertical-wrist diagram no longer applies (ABB 3HAC087213-001 Rev. H 2026, §1.6.1 p. 38 and §1.6.2 pp. 40 and 43; ABB 3HAC052852-001 Rev. U 2026, Revisions p. 7; FANUC RM-950iA catalog 2023, p. 2; FANUC RM-1000iA catalog 2022, p. 2; FANUC RM-2000 catalog 2024, p. 2).

How do axis speeds change across the heavy tiers?​

The heaviest arms compared here are markedly slower on the base axis. FANUC's M-2000 catalog gives a J1 maximum speed of 45°/s for the 1200F and 900F models and 20°/s for the 2300F and 1700F models, so the heavier pair takes more than twice as long to swing through the same base angle at full speed. Yaskawa's MH900 sheet gives maximum speeds of 45, 30, 30, 36, 37, and 70°/s for its S, L, U, R, B, and T axes.

For a cycle-time estimate, the arithmetic is simple and is UTEC's own illustration, not a manufacturer figure: a 90° base swing at 45°/s takes 2.0 s at full speed, and at 20°/s takes 4.5 s, before acceleration, deceleration, and settling time are added. The manufacturer's speeds are maxima, so acceleration, deceleration, and settling lengthen a real move beyond this estimate.

The comparison point is that choosing the 2,300 kg arm for a 1,100 kg load the 1200F model could carry buys payload margin at the cost of more than half the base-axis speed. A failure mode is a cycle-time promise built from maximum axis speeds with no allowance for the ramp times a heavy load requires (FANUC RM-2000 catalog 2024, p. 2; Yaskawa America DS-924 2021, p. 2).

What applications and construction details do the heavy-payload catalogs name?​

The catalogs name example tasks and construction details that help match a model to an industry, although none of them is a guarantee of fitness for a given job:

  • FANUC M-1000iA: the catalog names EV battery units and long building structures as example workpieces, says its high-rigidity design allows drilling, and notes that the J3 arm can stand upright and rotate behind the robot.
  • FANUC M-950iA/500: the catalog names friction stir welding, drilling, and riveting, and lists an accuracy and stiffness enhancement option.
  • FANUC M-2000: the wrist is rated IP67-equivalent.
  • Yaskawa MH900: the sheet gives a parallel-link construction, an IP67 wrist and IP30 body, and a 35 kVA power rating.

For heavy industry the environment rating is compared as carefully as the payload. The MH900's body is rated IP30 against its wrist's IP67, so for washdown in a pulp or paper mill, or spray in a steel-plant cooling area, the body rating, not the wrist rating, is the figure to compare with the area. A failure mode is choosing an arm on payload and reach for a wet or dusty area and discovering the body rating at installation (FANUC RM-1000iA catalog 2022, p. 2; FANUC RM-950iA catalog 2023, p. 2; FANUC RM-2000 catalog 2024, p. 2; Yaskawa America DS-924 2021, p. 2).

What does robot mass mean for the base, the foundation, and the cell structure?​

The published masses range from 2,500 kg for the M-950iA/500 to 12,500 kg for the M-2000/1700F-47A, and the ratio of robot mass to payload varies widely. Dividing the printed mass by the printed payload, an illustration computed for this article rather than a manufacturer figure, gives:

  • M-2000/2300F-37A: 11,000 ÷ 2,300 = 4.8
  • M-950iA/500: 2,500 ÷ 500 = 5.0
  • M-1000iA: 5,300 ÷ 1,000 = 5.3
  • IRB 8700-800/3.50: 4,750 ÷ 800 = 5.9
  • M-2000/900F-47A: 9,600 ÷ 900 = 10.7
  • MH900: 10,000 ÷ 900 = 11.1

In this list, the two 900 kg models with 4,683 mm reach carry the most robot mass per kilogram of payload. The mass figures, with the ABB figure excluding the dress pack and the FANUC figures excluding the controller, are the starting point for the foundation and any riser or pedestal; the load in motion adds reactions on top of the robot's weight, and those are taken from the manufacturer's foundation data for the chosen model.

A robot riser is a welded structure whose mounting face must stay flat for the robot to sit true. UTEC Industrial fabricates such bases and tooling frames, stress-relieves them with automated vibratory stress relief (VSR), and machines their mounting faces before assembly. The failure mode is a pedestal designed from the robot's static weight alone (ABB 3HAC087213-001 Rev. H 2026, §1.4 p. 23; FANUC RM-2000 catalog 2024, p. 2; FANUC RM-1000iA catalog 2022, p. 2; FANUC RM-950iA catalog 2023, p. 2; Yaskawa America DS-924 2021, p. 2).

What controls and sensing does a heavy-payload robot cell need, whatever the make?​

Each make compared here has its own robot controller, and the cell around it needs the same intelligence layer whichever make is chosen:

  • Cell PLC. The plant PLC sequences the conveyors, cars, and positioners that feed the robot and exchanges part identity, permissives, and status with the robot controller, for example over EtherNet/IP. The article on connecting a FANUC robot to an Allen-Bradley PLC covers that link, its timing, and its handshakes, so they are not repeated here.
  • Load confirmation. Load cells or part-presence sensors confirm that the part picked is the one the cell expected, so the robot runs the load case it was sized for. ABB provides a load identification routine in the controller for determining the load parameters.
  • Vision. Where part position or part number varies, machine vision locates or identifies the part before the robot grips it.
  • Safety. The robot falls under ISO 10218-1:2025 and the cell under ISO 10218-2:2025, adopted in the US as ANSI/A3 R15.06-2025. The MH900 sheet, dated 2021, cites the 2012 edition of R15.06, and A3 describes R15.06-2025 as a revision of that edition, so a new cell is specified to the 2025 edition. ISO 13849-1:2023 specifies a methodology for the design and integration of safety-related parts of control systems that perform safety functions, and applies to high-demand and continuous modes of operation.
  • Monitoring. Trending cycle time, drive faults, and gripper cycle counts can show wear and drift before a heavy part is dropped.

UTEC Industrial integrates FANUC robotic cells, including vision, with a FANUC design and engineering partner. A failure mode that does not depend on the make is a specification that names the robot but leaves the safety interfaces between the robot controller and the cell PLC unowned (ABB 3HAC087213-001 Rev. H 2026, §1.6.1 p. 38; Yaskawa America DS-924 2021, p. 2; ISO 10218-1:2025; ISO 10218-2:2025; ANSI/A3 R15.06-2025; ISO 13849-1:2023).

How should a buyer shortlist heavy-payload robots for a specific load?​

A shortlist is built from the load case, not from the brochure payload. A practical order is:

  1. Define the load. Part plus gripper mass, CG position from the flange, and the load's own inertia, for the heaviest part in the family.
  2. Filter by payload at the needed reach. Keep models whose printed payload covers the load at the reach the farthest pose requires, with the orientation each pose needs.
  3. Check the wrist. For FANUC and Yaskawa candidates, compare the load's moment and inertia with the printed ratings; for an ABB candidate, run the load diagram and inertia formulas.
  4. Check the environment and arm loads. Compare the IP ratings with the area, and place valve banks and dress packs within the arm allowances.
  5. Check cycle time and base. Estimate cycle time from the axis speeds with ramp allowance, and size the foundation from the robot's mass and foundation data.
  6. Compare accuracy on its label. Read each figure with its printed label and basis.

As an illustration, which is UTEC's own and not a manufacturer recommendation, take a 1,000 kg load that must be placed 3.2 m from the robot base with the wrist pointing down. On payload and reach alone, the M-1000iA's 1,000 kg at 3,253 mm covers it at full rating. The IRB 8700-800/3.50 covers it only through the vertical-wrist diagram, at that diagram's 1,000 kg maximum. The MH900's 900 kg rating does not cover it. The M-2000/1200F covers it with margin, at 8,600 kg of robot. The final choice then rests on the wrist check, the cycle time, and the base (FANUC RM-1000iA catalog 2022, p. 2; ABB 3HAC087213-001 Rev. H 2026, §1.6.2 p. 43; Yaskawa America DS-924 2021, p. 2; FANUC RM-2000 catalog 2024, p. 2).

Related Articles

References​

  • FANUC RM-950iA(E)-01: FANUC Robot M-950iA/500. FANUC Corporation, 2023.
  • FANUC RM-1000iA(E)-02: FANUC Robot M-1000iA. FANUC Corporation, 2022.
  • FANUC RM-2000(E)-01: FANUC Robot M-2000. FANUC Corporation, 2024.
  • ABB 3HAC087213-001 Rev. H: Product specification - IRB 8700 (OmniCore). ABB Robotics, 2026.
  • ABB 3HAC052852-001 Rev. U: Product specification - IRB 8700 (IRC5). ABB Robotics, 2026.
  • Yaskawa America DS-924: MH900: Large Reach, Heavy Payload Capacity. Yaskawa America, Inc., 2021.
  • ISO 9283:1998: Manipulating industrial robots — Performance criteria and related test methods. ISO, 1998.
  • ISO 9946:1999: Manipulating industrial robots — Presentation of characteristics. ISO, 1999.
  • Dagalakis, N. G. (1999). "Industrial Robotics Standards." In S. Y. Nof (ed.), Handbook of Industrial Robotics, 2nd ed. Wiley, pp. 447-459.
  • ISO 10218-1:2025: Robotics — Safety requirements — Part 1: Industrial robots. ISO, 2025.
  • ISO 10218-2:2025: Robotics — Safety requirements — Part 2: Industrial robot applications and robot cells. ISO, 2025.
  • ANSI/A3 R15.06-2025: American National Standard for Industrial Robots and Robot Systems – Safety Requirements. A3/ANSI, 2025.
  • ISO 13849-1:2023: Safety of machinery — Safety-related parts of control systems — Part 1: General principles for design. International Organization for Standardization, 2023.

Ready to Discuss a Material Handling System?​

UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for heavy industry, from the stress-relieved structure and drives to the Allen-Bradley PLC controls, tuning, and monitoring that run them, at its Spokane Valley, WA facility. Send UTEC the application, loads, and duty cycle to start a system review.

Request a Quote →

Questions? Call (509) 922-1832 or email sales@utec.co

Last updated on