Will a Cobot Stop Hands Near Presses at a Maryland Molding Plant?

Will a cobot stop when a hand gets near a molding press?

A cobot can be set to stop or slow when a person comes into its space, and that is one layer of protection. It protects people only when the whole cell has been designed and validated for the hazards that are actually there. In a Maryland molding plant the press is the machine that closes on hands, and the robot is a separate machine with its own motion path, so each one needs its own review.

OSHA has no robotics-specific standard. Its robotics page points instead to consensus standards: ISO 10218-1 and ISO 10218-2, with U.S. adoption as ANSI/RIA R15.06-2012, and ISO/TS 15066 for collaborative operation, with U.S. adoption as TR 606. Those are the documents your safety auditor and any integrator should be working from, and you should ask which edition each one uses.

Do not treat the word collaborative as proof of safety. Collaborative operation limits how a robot may contact people, but a molding cell still needs a documented risk assessment that covers the press, the robot, the part handoff, and every route a person might take into the area during a jam or a mold change.

How do power and force limiting and speed and separation monitoring work?

Power and force limiting means the robot stops when contact force exceeds body-region limits set in ISO/TS 15066. Speed and separation monitoring means the robot slows as a person gets closer and stops when the gap falls below a set minimum. Secondary guides describe both, but they do not reproduce the standard text, so check every number against the standard before it goes into a specification.

Neither mode tells you the cell is safe next to a press. Force limits assume the robot is the only moving hazard. In a molding cell, the press can close while the robot is reaching for a part, and a sensor placed to detect a person near the robot may not see the platen or the gap between mold halves at all.

Ask the integrator for four things in writing: the protective stop behavior, the sensors and their safety rating, the speed and separation values used with the reasons for them, and the test records for each place where the robot and a person could meet. A vendor who cannot produce these is not ready to lease you a cell.

Why is the molding press the bigger hazard?

A closing mold can amputate a hand in a single cycle. A guide on injection molding safety names the operator gate and its interlocks as the primary defense, and it points to 29 CFR 1910.147 for energy control during servicing. For the machine standard it cites ANSI/PLASTICS B151.1-2017, but an older OSHA-archived list shows the 1997 version, so confirm which revision applies before you write anything down.

In our reading of the sources, injection molding machines are generally covered by the general machine guarding rule, 29 CFR 1910.212, rather than the mechanical power press rule, 29 CFR 1910.217. That is an interpretation, not a stated OSHA position, so confirm the classification with your counsel or your OSHA area office before you finalize procedures.

Guards are often defeated in practice, and that is a common failure the guide flags. OSHA's machine guarding hazards page states that replacement guard windows must meet or exceed the original manufacturer specifications. A cobot cell that adds a second gate or a bypass makes the press harder to protect, so every guard change should be logged and rechecked.

Mold changes deserve their own review. Changeovers often mean a guard is open, the robot is repositioned, and the press is in a state nobody modeled during the original risk assessment. Build a changeover step into the cell's procedure and into its validation test.

What do injury rates say about plastics plants?

BLS reports that plastics and rubber products manufacturing (NAICS 326) had a 2024 total recordable case rate of 2.8 per 100 full-time workers, and a rate of 1.8 for cases with days away, restriction, or transfer. The 2023 rate was 2.9, but BLS notes a series break from NAICS 2022, so those years may not be directly comparable.

An ISHN article states that machinery operators and maintainers suffer about 18,000 amputations, lacerations, crushing injuries, and abrasions per year, and over 800 deaths. The excerpt we reviewed does not name the underlying source, so cite it as the article's own figure rather than an official count.

Sector rates are context for your safety committee, not a measure of your plant. Your own OSHA 300 log is the baseline to compare against after any change, and it is the figure that tells you whether a change actually helped.

Can a cobot take over the hand reach into a press?

Sometimes. A cobot can remove parts or sprues through a fixed opening, which takes a regular reach out of the press area. It should not be sent into the closed press space, and it cannot judge a jam the way a trained person can. The comparison below sets the two side by side.

Where a person still has to enter the press area, the procedure does not change because a robot is present. Energy has to be isolated and verified, the press cannot cycle, and the person working inside must control the lockout under the energy control rule. A cobot can shrink how often that happens, but it cannot remove the need for the procedure, and any plan that relies on the robot to stop the press in time has not been validated.

What does the Maryland labor market mean for this decision?

Molding machine setters, operators, and tenders (SOC 51-4072) are a large occupation. BLS OEWS lists about 158,980 people in the group, though the year is not stated in the excerpt we reviewed, so treat that figure as a national reference and check the table for the year you need. The notes did not include a turnover rate for this job, which would need a separate source such as JOLTS.

Maryland's manufacturing base is shifting. According to FRED, manufacturing employment in the Baltimore-Columbia-Towson metro was about 56.1 thousand in December 2025, down from 57.1 thousand in August 2025. A Daily Record report says manufacturing lost about 1,000 jobs in April 2026 in the state's monthly estimate.

Falling employment does not show whether you can fill your own openings. Look at your own posting history, overtime hours, and turnover on the press floor, because those numbers decide whether a cobot relieves a shortage or only changes its shape. Staffing pressure is a reason to look at a cell, but it is not a reason to skip press guarding.

What should a safety manager check before deciding?

Before you request a quote, gather the facts that can change the answer. Most of them live inside your plant, and none of them come from a brochure.

If the press guarding is sound, the jam procedure is locked out, and a clear part path exists, a cobot cell may be a reasonable next step. If any of those three is missing, fix it first, because no robot will compensate for an open gate.

Ask who maintains the cell after go-live and how quickly a fault gets a response. A cell that stops often for sensor faults will tempt operators to bypass it, and a bypassed safety function is worse than no cell at all. Ask the integrator how they handle that pressure during training.

How does leasing and deploying a cell work with us?

We are an OEM-neutral integrator, so we choose the robot that fits your press and floor rather than whatever one manufacturer offers. We then lease it through financing, deploy and integrate it, train your operators and maintenance staff, and service it under our service plan through regional service engineers in every state we service. One partner handles the whole lifecycle.

The first step is a free site assessment that maps the press cell, the part flow, and the guarding, and it returns the safety questions in writing. The lease payment comes in a quote after that assessment, not before. If the assessment finds the press guarding needs work, we say so first, because a cobot placed behind an unguarded press does not solve the problem.

Most plants begin with one leased cell on one press, run a phased deployment that keeps the rest of the floor producing, and expand only after validation results and the first months of records support it. Timing depends on the guarding and integration work, and we confirm it in the site assessment rather than promising a date here.