Skip to main content

When Does a Robot Beat a Custom Mechanism for Heavy Material Handling?

A robot beats a custom mechanism for heavy material handling when the plant has to move many different parts along changing paths, and a custom mechanism beats a robot when one heavy part family follows one fixed path at high load. 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 out the decision from the customer's side: part variety, load and moment, the safety standards that apply to each option, the controls and sensing each one needs, and how each is tuned and monitored once it runs. Either answer is built along the same chain, design → engineering → parts machining → fabrication → assembly → weld fatigue → stress relief → drives → controls → tuning → monitoring, and the choice made at the design link sets what the controls and monitoring links have to do.

What decides whether a robot or a custom mechanism should move a heavy part?​

As engineering reasoning, the decision turns on how much the motion has to change, not on how modern the equipment looks. An industrial robot is a general-purpose manipulator: its value is that the same arm can follow a new path, pick a new part, or serve a new machine after a program change. A custom mechanism, such as a transfer car, a lift-and-carry beam, a tilting positioner, or a dedicated manipulator, is built for one load path and can do that path with less hardware, less programming, and more structural margin.

As engineering practice, the comparison comes down to a few questions:

  • How many part families? One casting pattern moving from a shakeout line to a cleaning room favors a mechanism; a mix of forgings, weldments, and machined parts serving several stations favors a robot.
  • How fixed is the path? A fixed transfer between two known points is a mechanism's strength; a path that changes by part number, stack height, or pallet pattern is a robot's.
  • How heavy is the load relative to the rated payload and moment of available arms? A load that sits near or beyond an arm's rated limits pushes the answer to a mechanism.
  • What is the environment? Radiant heat at a furnace door, abrasive dust in a mineral plant, or washdown in a paper mill affects both options, but a robot's wrist, cabling, and dress pack can be the most exposed parts.
  • Who maintains it? A plant with PLC and drive technicians but no robot programmers carries a real support cost for a robot.

The safety framework differs too: the robot-level standard addresses a robot as partly completed machinery with its own safety requirements, while the robot application and cell are covered by a separate cell-level standard, so, as engineering reasoning, choosing a robot also selects the safety standard set that the design must follow (ISO 10218-1:2025; ISO 10218-2:2025).

How does part variety change the robot-versus-mechanism decision?​

As engineering reasoning, part variety can tip the answer. A new part family added to a custom mechanism can mean new tooling, new fixtures, or a new mechanical axis; a new part family added to a robot cell can mean a new program and, where the grip changes, new end-of-arm tooling.

Consider three examples drawn from different industries:

  • An aluminum or steel plant moving one ingot or slab size from a furnace to a mill runs the same motion thousands of times a year. A roller table, a charging car, or a dedicated transfer is simpler to build and keep running than an arm.
  • An aerospace assembly line loading several fixture types onto machine tools, where the part mix changes with each program, is the kind of variable job a robot handles well.
  • A heavy fabrication shop unloading weldments of many sizes from a positioner can go either way: a robot if the parts stay within its payload, a mechanism with adjustable tooling if they do not.

As engineering practice, a useful rule of thumb is to count the distinct pick poses, drop poses, and grip types the system must handle over its first five years. When that count is low and stable, the mechanism tends to win on cost and uptime; when it is high or unknown, the robot's reprogrammability can be worth its higher integration effort; the payback method is covered in the robot cell ROI article. Whatever the count, the risk assessment for the cell should cover every task on that list, including tool changes and part changeovers, since the task-based method scores severity, exposure, and avoidance for each task (RIA TR R15.306-2016).

When do payload, moment, and reach rule a robot out?​

A robot's limits are set by its manufacturer's rated payload, allowable wrist moment and inertia, and reach, and a heavy part can exceed the moment limit long before it exceeds the payload limit. The moment is the part's weight multiplied by the distance from the wrist flange to the part's center of gravity, so a long or offset part is harder on a robot than its weight suggests.

A simple calculation shows why. Take a 1,000 kg part whose center of gravity sits 0.5 m from the wrist flange once the gripper is attached:

  • Weight: W = m × g = 1,000 kg × 9.81 m/s² = 9,810 N
  • Static wrist moment: M = W × d = 9,810 N × 0.5 m = 4,905 N·m

If the gripper's geometry doubles that offset to 1.0 m, the moment doubles to 9,810 N·m with no change in the part's weight. Acceleration during a fast move adds dynamic load on top of the static figure. The specifying engineer compares these numbers, including the gripper's own mass and offset, with the manufacturer's published limits for the arm; a load that sits at or beyond them points to a mechanism, a lift assist, or a robot that only guides a load carried by a separate axis.

As engineering reasoning, other conditions also rule a robot out or force a hybrid design:

  • The part must be held for a long dwell, such as in a quench tank or a curing station, where a mechanism holds it at far lower cost.
  • The load path runs through a furnace door or a hot zone that the arm's cabling and seals cannot tolerate.
  • The move is long, such as bay-to-bay travel, which puts the robot on a track or leaves the travel to a car or crane.

As engineering reasoning, a failure mode to watch for in heavy cells is a gripper designed after the robot was chosen: the added tooling mass and offset push the arm past its moment rating, and the cell is re-scoped late in the project. The safety requirements for the robot itself, as distinct from the cell around it, sit in the robot-level standard, which addresses robots as partly completed machinery (ISO 10218-1:2025).

Are robots now common outside automotive plants?​

Yes. The Association for Advancing Automation reported that North American robot orders reached 36,766 units worth $2.25 billion in 2025, a 6.6 percent increase in units ordered over 2024 on a consistent reporting cohort, and that demand from non-automotive customers outpaced demand from automotive customers, with general industries capturing the majority share of units ordered. The industries A3 names are food and consumer goods, semiconductors and electronics, and life sciences, not heavy industry.

The rest of this answer is engineering reasoning, not A3 data. Robots are no longer an automotive-only tool, which matters for heavy-industry buyers who may assume the technology and its support base sit in car plants. For a plant engineer, that shift suggests three practical effects:

  • Integrators, spare-parts stock, and trained technicians can be found outside automotive supply chains.
  • Robot cells can be specified by plants that have no robot standards or programmers in-house, which raises the importance of a written risk assessment and a clear handover.
  • Buyers can expect to compare robot quotes against custom-mechanism quotes, so the decision factors in this article can come up in the same bid review.

The market data does not settle the decision for a given heavy part; it only shows that robots are an option to evaluate alongside a mechanism (Association for Advancing Automation 2026).

How do safety requirements differ between a robot cell and a custom mechanism?​

A robot cell and a custom mechanism fall under different standard sets, and the difference affects scope, documentation, and cost. For a robot, ISO 10218-1:2025 covers safety requirements for the industrial robot itself, addressed as partly completed machinery, and ISO 10218-2:2025 covers the safety of the robot application and the robot cell, including integration, commissioning, and operation. The iso.org FAQ for the 2025 edition of Part 2 says it "complements standards such as ISO/TS 15066", the collaborative-robot technical specification (ISO/TS 15066:2016). In the United States, ANSI/A3 R15.06-2025 is the national adoption of ISO 10218 Parts 1 and 2, so a US robot cell can be specified to R15.06-2025.

A custom mechanism has no robot-specific standard. Its safety case is built from general machinery standards:

  • ISO 12100:2010, which specifies principles of risk assessment and risk reduction for machinery design
  • ISO 13849-1:2023, which covers the design and integration of the safety-related parts of control systems that perform safety functions, in high demand and continuous modes of operation
  • IEC 60204-1:2016+AMD1:2021, which applies to the electrical equipment of machines not portable by hand while working

The rest of this paragraph is engineering reasoning. These general standards also apply to the robot cell alongside the robot standards. The practical consequence is that a robot cell carries two layers of safety documentation, the robot manufacturer's for the arm and the integrator's for the cell, while a mechanism carries one integrated layer from its builder. A failure mode to design out in either case is a safety boundary with no owner: the robot is certified, the conveyor is certified, and the interface between them, such as a shared gate or a part hand-off, is covered by neither. UTEC Industrial integrates FANUC robotic cells, including vision, with a FANUC design and engineering partner (ISO 10218-1:2025; ISO 10218-2:2025; ANSI/A3 R15.06-2025; ISO 12100:2010; ISO 13849-1:2023; IEC 60204-1:2016+AMD1:2021).

How is a heavy-handling robot cell risk-assessed?​

One US method for robot cells is the task-based risk assessment described in RIA TR R15.306-2016, which scores each hazard to a person performing a task in or near the cell on three factors:

  • Severity of injury: S1 to S3
  • Exposure, meaning the estimated frequency or duration of exposure: E0 to E2
  • Avoidance, meaning a person's ability to sense and elude the hazardous situation: A1 to A3

The combination sets the initial risk level, and risk-reduction measures are then selected and the residual risk determined. The rest of this answer is engineering reasoning. The task-based approach matters more for heavy handling than for light assembly, for three reasons:

  • Severity starts high. A heavy part dropped or swung by a robot, or a mechanism that pinches a person against a fixture, is likely to score at the top of the severity scale.
  • Maintenance and changeover can dominate exposure. Operators may seldom stand next to a running heavy cell, but maintenance technicians clearing a jammed part, changing a gripper, or teaching a new position may enter it daily.
  • Avoidance can be poor. A person cannot step clear of a 1,000 kg part moving at robot speed.

The same method can be applied to a custom mechanism, which makes it a fair basis for comparison: list the tasks for each option and compare the residual risk after safeguarding. The task list, not the equipment type, drives how much guarding, safety-rated sensing, and procedure the system needs. A failure mode to guard against is a risk assessment that lists only production tasks and omits teaching, recovery from a dropped part, and cleaning, which are tasks that put people inside the cell (RIA TR R15.306-2016; ISO 12100:2010).

How far must guarding sit from a heavy robot or mechanism?​

Where a light curtain, laser scanner, or other presence-sensing device stops the hazard, the device must sit far enough away that a person cannot reach the hazard before it stops. ISO 13855:2024 specifies requirements for positioning and dimensioning safeguards with respect to the approach of the human body or its parts towards hazards, and its approach speeds (walking speed and upper limb movement) are used to derive separation distances. Its scope excludes approaches such as running, jumping or falling, safeguards that can be moved by hand nearer the hazard zone than the separation distance, hazards from emissions, and failure of mechanical parts or gravity falls; it applies to protection of persons 14 years and older, and its separation distances do not apply to safeguards used solely for presence sensing.

The rest of this answer is engineering reasoning. The separation distance also depends on the overall stopping time, which is the sum of every delay between detection and standstill:

  • The sensing device's response time
  • The response time of the safety-related control system, including the time a safety signal takes to cross the network
  • The drive's or brake's response time
  • The time for the robot or mechanism to come to rest

Heavy loads lengthen the last term: a robot carrying a heavy part, or a mechanism moving a heavy car, takes longer to stop, and a longer stopping time pushes the guard farther out. A failure mode is a guard distance calculated from the robot manufacturer's stopping data for an empty arm, then applied to the arm carrying its heaviest part (ISO 13855:2024; ODVA PUB00138R8-2024).

What controls and sensing does a heavy-handling robot or mechanism need?​

As engineering reasoning, both options need the same intelligence layer: position feedback, a controller that sequences each move and enforces permissives, safety-rated stopping, and drives matched to the load. What differs is where the motion control lives. In a robot cell, the robot controller owns the arm's motion and the plant PLC coordinates the cell; in a mechanism, the PLC and its drives own the motion directly.

The layers are below; the sensing bullet is engineering practice, and the others follow the cited documents:

  • Network. The robot can connect to the PLC over EtherNet/IP. ODVA classes a robot that sends and receives real-time data at a PLC's request as an Adapter Class device, and the PLC as the Scanner Class device that originates the I/O connection. Implicit I/O data travels over UDP/IP at regular intervals, while explicit messages, generally used for configuration and diagnostic data, use TCP/IP.
  • PLC task structure. Logix 5000 controller tasks can be configured as continuous, periodic, or event, and a periodic task executes at a constant period, so cell interlocks and hand-off logic can run at a fixed period.
  • Safety logic. Safety functions such as the emergency stop and guard interlocks can be realized by safety-related parts of control systems designed to ISO 13849-1:2023, for high demand and continuous modes of operation.
  • Drives. In a mechanism, a servo axis can be driven by an Allen-Bradley Kinetix 5700 inverter. The manual describes the 2198-xxxx-ERS3 and -ERS4 dual-axis and single-axis inverters as equipped for hardwired safe torque-off and for integrated safe torque-off over the EtherNet/IP network, and the inverters support DSL and Hiperface encoder feedback; the manual also states that safe torque-off does not provide physical isolation of the electrical output that is required for some applications. A VFD ramps a car or conveyor.
  • Sensing. Encoders and limit switches report position; load cells confirm the weight actually picked; part-presence sensors and, where parts vary, machine vision confirm what the robot is about to grip.
  • Electrical basis. IEC 60204-1:2016+AMD1:2021 applies to the electrical equipment of machines not portable by hand while working, starting at the point of connection of the supply.

As engineering reasoning, a failure mode specific to robot cells is safety handled twice with no clear owner: the robot's internal safety functions and the PLC's safety logic each assume the other stops the shared conveyor. UTEC Industrial, a Rockwell Automation Recognized System Integrator, builds this Allen-Bradley PLC, drive, and safety layer into the handling systems it delivers (ODVA PUB00138R8-2024; Rockwell Automation 1756-RM094N-EN-P-2025, Ch. 5 pp. 39 and 41; Rockwell Automation 2198-UM002E-EN-P, Kinetix 5700, pp. 16, 241 and 250; ISO 13849-1:2023; IEC 60204-1:2016+AMD1:2021).

How is stored energy controlled when a robot or mechanism holds a heavy load?​

A heavy load held in the air is stored energy, whether a robot gripper or a mechanism's lift axis holds it. Stopping the program does not remove that energy. OSHA's lockout/tagout rule, 29 CFR 1910.147, covers servicing and maintenance of machines in which unexpected energization, start up, or release of stored energy could cause injury. It defines an energy-isolating device as a mechanical device that physically prevents the transmission or release of energy and states that push buttons, selector switches, and other control-circuit-type devices are not energy-isolating devices. Paragraph 1910.147(d)(5)(i) requires all potentially hazardous stored or residual energy to be relieved, disconnected, restrained, and otherwise rendered safe after lockout or tagout devices are applied, and 1910.147(d)(6) requires verification of isolation and de-energization before work begins.

For heavy handling, that has design consequences on both sides; the first two bullets are engineering reasoning:

  • Robot cells. A part left in the gripper when power is removed stays there only as long as the gripper's clamping holds. Vacuum and pneumatic grippers need a defined state on loss of air, and the procedure must say how the part is lowered or supported before anyone enters.
  • Mechanisms. A raised lift table, a tilted positioner, or a car parked on a slope stores gravity energy that needs mechanical blocking, pins, or a lowered state before work begins.
  • New equipment. Paragraph 1910.147(c)(2)(iii) requires that, after January 2, 1990, whenever new machines or equipment are installed, or replacement or major repair, renovation, or modification of a machine is performed, its energy-isolating devices be designed to accept a lockout device.

As engineering reasoning, a failure mode to design out is a technician entering a stopped cell to free a jammed part while the part is still held only by a gripper or brake (OSHA 29 CFR 1910.147-1989).

How are robot and mechanism cells tuned and monitored after startup?​

As engineering reasoning, settings that worked at startup can drift, and heavy loads make drift expensive. Tuning and monitoring, the last two links of the chain, keep the system's limits real:

  • Servo tuning. The Kinetix 5700 commissioning procedure includes a Tune the Axes step, with the procedure chosen by motor type. For permanent-magnet motors the manual says a load observer with default gains means that most of the time no auto-tune is needed; for induction motors it directs tuning with the load removed first, then re-attaching the load and tuning again to provide an accurate operational response. As engineering reasoning, a mechanism tuned empty can overshoot once a heavy part is clamped.
  • Network timing. Rockwell Automation's guideline is to specify an I/O RPI at 50 percent of the rate at which new data is needed, which its example shows is half the update interval: if information is needed every 80 ms, the RPI is set at 40 ms, so the data is sampled twice as often. Because I/O updates are asynchronous to the logic scan, an input can change mid-scan unless it is buffered.
  • Condition monitoring (engineering practice). Trending drive current, fault history, gripper cycle counts, and cycle times shows wear and slow drift before a dropped part or a stalled axis.

As engineering reasoning, for a robot the robot controller's own diagnostics add to this picture; for a mechanism, the PLC and drives carry all of it. Either way, a monitored cell tells the plant when a tuning or timing change is needed instead of waiting for a fault (Rockwell Automation 2198-UM002E-EN-P, Kinetix 5700, pp. 207 and 210; Rockwell Automation 1756-RM094N-EN-P-2025).

What should a specifying engineer define before choosing a robot or a mechanism?​

A request for quotation that says only "robot" or only "mechanism" invites two bids that cannot be compared. As engineering practice, a specification that lets both options be evaluated on the same basis defines:

  • Part list: every part family, weight, dimensions, center-of-gravity location, and grip or support points, for the first five years.
  • Motions: pick and drop poses, travel distance, required cycle time, and dwell times.
  • Environment: temperature at pickup, dust, coolant, washdown, and any hazardous-location classification.
  • Tasks: every human task in or near the equipment, including teaching, changeover, maintenance, and recovery from a dropped or jammed part, because the task-based risk assessment scores each task separately.
  • Safety scope: which standard set applies, who owns the risk assessment for each interface, and the required safety performance of each safety function.
  • Controls: PLC platform, network, and how the robot or mechanism exchanges standard and safety signals with the plant.
  • Support: who will program, tune, and maintain the equipment after handover.

With those inputs, a robot bid and a mechanism bid can be compared on residual risk, cycle time, and life-cycle cost rather than on first price. The task list and interfaces defined here are the same inputs the integrator's risk assessment for the cell will need (RIA TR R15.306-2016; ISO 10218-2:2025).

Related Articles

References​

  • 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.
  • RIA TR R15.306-2016: Task-Based Risk Assessment Methodology. Robotic Industries Association/A3, 2016.
  • Association for Advancing Automation (2026). Robot Orders Grow 6.6% in 2025 as General Industries Drive Broader Automation Adoption. Press release, A3, 2026.
  • ISO 12100:2010: Safety of machinery — General principles for design — Risk assessment and risk reduction. ISO, 2010.
  • ISO 13849-1:2023: Safety of machinery — Safety-related parts of control systems — Part 1: General principles for design. International Organization for Standardization, 2023.
  • IEC 60204-1:2016+AMD1:2021 CSV (Ed. 6.1): Safety of Machinery -- Electrical Equipment of Machines -- Part 1: General Requirements. International Electrotechnical Commission, 2021.
  • ISO 13855:2024: Safety of machinery — Positioning of safeguards with respect to the approach of the human body. ISO, 2024.
  • ODVA PUB00138R8-2024: EtherNet/IP — CIP on Ethernet Technology. ODVA, 2024.
  • Rockwell Automation 1756-RM094N-EN-P-2025: Logix 5000 Controllers Design Considerations. Rockwell Automation, 2025.
  • Rockwell Automation 2198-UM002E-EN-P (2018): Kinetix 5700 Servo Drives User Manual. Rockwell Automation, 2018.
  • OSHA 29 CFR 1910.147-1989: The Control of Hazardous Energy (Lockout/Tagout). Occupational Safety and Health Administration, 1989.

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