Can a Cobot Take the Foam Core Wrap Station at a Mattress Plant?

Can a cobot take the foam core wrap station at a mattress plant?

Sometimes, and the test is more specific than whether a robot can move like a wrapper. A cobot is a collaborative robot arm built to share space with people, and it can take over a repeated wrap motion at a single station when the foam cores arrive in a predictable range, the wrap pattern is fixed, and the cell leaves room for the arm and a clear path for the operator. The harder question is whether the arm can hold the station's cycle time while cutting the strain that is pulling people off the line.

The sector around your station is shrinking in business count. According to a 2025 IBISWorld industry report, about 314 mattress manufacturing businesses operated in the US in 2025, and that count declined at a 0.7 percent compound annual rate from 2020 to 2025. For your plant, the sector figure matters less than your own station, where the decision turns on cycle time, fixturing, floor space and how the wrap work is staffed.

Your situation has two linked problems. The first is the physical strain on the people who wrap by hand. The second is the gap that rotation leaves when a person comes off the station. A cobot can address the first only if it takes enough of the repetitive motion, and it can address the second only if the line keeps moving while the arm works.

Why does hand wrapping wear out shoulders?

Repeated reach, lift and wrap work is one of the most common paths to shoulder injury, and federal data show how large that problem is. According to the U.S. Bureau of Labor Statistics, shoulder injuries kept workers away a median of 18 days across private industry, and about half came from overexertion. These figures come from a BLS release dated 2007 in our research, so treat the exact values as history, but the mechanism still matches a wrap station.

Manufacturing carries a heavy share of these injuries. In the BLS manufacturing summary dated 2003 in our research, the sector logged 225,800 lost-worktime cases, and 82,160 of them, more than a third, were musculoskeletal disorders. MSD cases averaged a 13-day absence, and the shoulder alone accounted for 16,010 lost-worktime cases.

More recent federal data point the same way. According to the BLS MSD fact sheet, which uses 2018 data, retail trade, manufacturing, and healthcare and social assistance together had 50 percent of all private-sector MSD cases. Repetitive manual work in plants is a national pattern, not a quirk of one building.

Enforcement records show how fast a single line can go wrong. In a 2016 OSHA case, an auto-seat plant recorded six surgical musculoskeletal injuries in 2014, 15 in 2015 and one by mid-2016, with a lost work time rate more than twice the BLS industry average. A shoulder problem at one station is easy to undercount until the absence records show it.

How much shoulder load does a cobot actually take off?

The evidence is encouraging but limited. A 2019 IEEE study found that an optimized human-robot collaboration setup cut shoulder joint torque overloading by 8.24 Nm, plus or minus 0.97, compared with initial configurations, with p under 0.001. That is a measured effect in a studied setup, not a promise about your wrap station.

A closer match to manual handling comes from a HAL and LIRMM study with 11 participants. Right shoulder humeral rotation fell from 68.4 percent to 7.4 percent of maximum voluntary contraction when a cobot assisted the handling task. The drop is large, but the sample is small and the task is not your wrap motion.

The limit is also clear. A 2025 arXiv preprint, which has not been peer reviewed, found that even optimized joint torques still averaged roughly 15 Nm in the shoulder and 12 Nm in the elbow. Assistance lowers shoulder load without removing it.

The practical read: a cobot is a way to cut peak strain at one station, and it works best alongside station height, part presentation and job rotation, not in place of them.

Will a cobot slow the wrap station down?

Only if the cell is built that way. The test is the station's cycle time against the rate the line needs, measured with your actual foam cores, wrap material and fixtures in place. A slower but steady cobot cycle can still keep the line moving if the stations before and after it are balanced to that rate. A cycle that cannot hit the rate will leave downstream people waiting.

Watch for the time around the wrap. Part presentation, pickup, the handoff to the next station and the reset after a fault often take longer than the wrap itself. An operator absorbs some of that time into natural movement, while a robot shows it plainly, so a time study on one shift before the pilot tells you more than the arm's spec sheet.

Feeding and staging matter too. If foam cores arrive in batches or still get staged by hand, the arm waits on people and the line slows. A buffer or a simple stop on the conveyor can cover short gaps, but size it from your own data, not from a guess.

What does a shoulder injury cost the line in lost days?

The direct cost is time away. The BLS release dated 2007 puts the median shoulder injury absence at 18 days across private industry, and the manufacturing summary dated 2003 puts the average MSD absence at 13 days. Those are not the latest medians for mattress makers, but they show how much one shoulder case removes from a station that has no spare hands. A shoulder case also pulls supervisor time into scheduling and retraining.

A station that loses one person for two weeks loses coverage, and the rotation that follows can push strain onto the people who cover it. This is where a cobot earns its place: it can keep the station running on the days a worker is out, or give the rotation time to recover, as long as a trained operator stays on the cell.

Our research could not confirm turnover for upholstery, foam or sewing workers, and it did not find a usable skills shortage figure for this job. Pull your own absence and turnover logs for the station before deciding. Those records will set the value of covering the station more than any national average will.

How common are cobots on factory floors?

Cobots are growing but still a minority of industrial robots. According to the International Federation of Robotics, new cobot installations reached 66,000 units in 2025, which was 11 percent of all industrial robot installations. Cobot growth of 7 percent fell below the overall industrial robot market for the first time, which suggests the category is maturing rather than exploding.

The IFR baseline is useful for context. In 2024, more than 64,500 cobots were sold, nearly 12 percent of installations, with 13 percent year-on-year growth. Totals are less reliable: one IFR page cites 603,000 industrial robot units, a figure that does not match the 354,000 reported in trade press, so this page does not rely on any total count.

For a plant, the practical point is that cobots are common enough that integrators should be able to show references on repetitive handling cells. Where a plant already runs cobots on other cells, operators and maintenance staff often know the controls, which can shorten training. Ask for those references, and ask whether the same team will service the unit after go live.

What safety rules apply when a cobot works next to people?

A cobot that shares a station with an operator falls under a risk assessment, and the main reference is ISO/TS 15066:2016, a Technical Specification that gives force and pressure limits by body region for human-robot contact. It defines four collaborative operation modes, including power and force limiting. Because it is a Technical Specification and subject to revision, confirm the latest edition with your integrator and your safety engineer.

ISO 10218 parts 1 and 2 set general requirements for industrial robots and their integration. We did not verify the edition status of those parts for this page, so do not treat any edition date here as settled.

Safety design shapes the layout. Guarding, speed limits near the operator and the spot where people load cores all decide how much of the station the arm can use. A cell that runs slowly near people can still be the right choice if its rate matches the line, but that has to be confirmed in a layout review, not assumed from a brochure.

Our site assessment maps the cell, the operator path and the arm's reach before any quote is written, so the layout review starts from measured conditions.

How do you get a cobot on the wrap station with us?

We are an integrator, not a single-brand seller. We choose the cobot for your wrap cell from more than one manufacturer, lease it through financing, deploy and integrate it, train your staff and service it. You work with one partner across the whole lifecycle, not with a separate vendor for each step.

The first step is a free site assessment. We map the station, measure the foam core and wrap ranges, time the steps with your team and check the operator path and reach. The lease payment is quoted after that assessment, so the number reflects your station, not a generic estimate.

We usually recommend starting with one leased unit and a phased deployment, so the line keeps running while the cell is tuned. Every unit we place is serviced under our service plan, with regional service engineers in the states we cover, and we can dispatch on site or triage remotely.

If the assessment shows that the cycle cannot hold the line rate, or that the cores vary too much, we will say so and recommend the fixture or ergonomic change that fits better.