The Design Review That Should Have Happened Before the First Cut

We had a machine come back from install with a problem nobody caught. The infeed track delivered the part to the first station, but the part sat 8 mm too low. The gripper couldn’t reach it. The fix was shimming the entire station up 8 mm on the customer’s floor. It took four hours and three technicians. The root cause? Nobody on the design review had actually stood at the operator’s position and looked at how the part arrived at the station.

Custom machine design review is not a meeting where the engineer presents the model and everyone nods. It’s a structured, item-by-item check that catches the things the drawer is blind to — because the drawer spent six weeks looking at this model and can no longer see what’s obvious to a fresh pair of eyes.

This is the checklist I bring to every design review. It’s grown over years of punch-list items, customer callbacks, and “why didn’t we think of that?” moments. It’s not exhaustive, but it covers the 90% of issues that actually show up on the floor.

1. Operator Access and Ergonomics

The operator is the person who uses this machine every day. If they can’t reach, load, unload, or service it comfortably, the machine is a failure — no matter how clever the mechanics are.

What to Check

  • Loading height: Is the part load/unload position between 700 mm and 1,100 mm from the floor? Below 700 mm and the operator is bending. Above 1,100 mm and they’re reaching. The sweet spot is 850–950 mm.
  • Reach into the station: Can the operator reach the part without leaning over a guard? If they have to reach more than 300 mm into the machine, they’ll stop doing it. Add a tray, a slide, or a closer station.
  • Visibility: Can the operator see the part at the critical station without craning? If the tooling is hidden behind a frame member, they can’t see when something goes wrong. Add a window or position the station in plain sight.
  • Emergency stop reach: Can the operator hit the e-stop from every operating position? If they have to let go of the part and walk two steps, it’s not reachable.
  • Weight of change parts: Can one person lift the changeover tooling? If it’s over 15 kg, it needs handles, a lift assist, or it needs to be split into two pieces.

2. Maintenance and Service Access

The machine will break. The question is whether the technician can fix it in ten minutes or four hours. The design review is where you decide which.

What to Check

  • Sensor replacement: Can every photo-eye, proximity switch, and pressure switch be replaced without removing tooling? If a sensor behind a gripper requires taking the gripper off, it’s bad design. Move the sensor or add a quick-mount bracket.
  • Cylinder replacement: Can a cylinder be removed without disconnecting the frame? If the mounting bolts require a socket extension from the back, it’s a four-hour job. Design for access.
  • Filter and lubricator: Is the FRL visible and accessible? If it’s buried inside the frame, nobody will drain the filter. Mount it where it’s at eye level and easy to reach.
  • Bearing and linear rail replacement: Are the wear items (bearings, rail carriages, belts) accessible with standard tools? If a linear rail carriage requires removing three other components to get to it, it’s not serviceable.
  • Clearance for tools: When a technician needs to tighten a bolt, is there room for a socket wrench? A bolt in a corner with 20 mm clearance for the head but no room for the wrench is a design error.

The 15-minute rule: If a maintenance task can’t be done in under 15 minutes with standard tools, it’s going to be done badly. Design for 15-minute service.

3. Safety and Pinch Points

Safety isn’t just about the guarding you add. It’s about the geometry of the machine itself. A pinch point between a moving gripper and a fixed frame is a hazard regardless of the guard around it.

What to Check

  • Pinch points: Walk the mechanism through its full motion. Where do two moving parts come together? Where does a moving part approach a fixed frame? Any gap between 6 mm and 25 mm is a finger trap. Widen it to under 6 mm or over 25 mm.
  • Crush points: Where does a heavy load (gripper, slide, rotary table) come to rest? If an operator’s hand could be in that space during normal operation, it needs a guard or an interlock.
  • E-stop categories: Does the e-stop stop motion in the right way? A vertical axis on e-stop should hold position (servo brake) or fall safely (controlled descent), not drop.
  • Guarding interlocks: Every removable guard panel has an interlock. No exceptions. A guard without an interlock is a suggestion, not a safety device.
  • Air bleed on e-stop: On e-stop, do vertical pneumatic cylinders drop? They shouldn’t. Pilot check valves hold them. This is a circuit check, not a mechanical one.

4. Part Flow and Tooling Access

The part has to get in, get through the machine, and get out. The design review is where you trace the part’s path and find the places where it gets stuck.

What to Check

  • Part infeed path: Trace the part from the feeder through the infeed to the first station. Does it have clearance at every transfer point? Does it hit a sensor, a bracket, or a frame member?
  • Part outfeed path: Trace the part from the last station to the outfeed. Can it fall clear? Does it get caught on a bolt head or a bracket edge?
  • Jammed part access: If a part jams at a station, can the operator clear it without disassembling the tooling? If they have to take the gripper off to remove a jammed part, the machine will be down for hours every time it jams.
  • Tooling clearance: When the tooling closes, is there clearance for the part’s maximum allowed variation? If the customer’s part has ±0.2 mm tolerance, the tooling needs to accommodate that without jamming. Run the max and min part through the model.
  • Dirty parts: If the part comes in with oil, flash, or packaging debris, where does it go? If it collects in the tooling or on a sensor, it’ll cause misreads. Add a drain, a brush, or an air blow-off.

5. Fasteners and Assembly

This is where the shop foreman catches things the designer misses. The designer knows the bolt size; the shop knows how long it takes to tighten a bolt you can barely reach.

What to Check

  • Fastener standardization: Are all bolts the same thread (e.g., all M8) wherever possible? A machine that uses M5, M6, M8, and M10 bolts in the same area requires four sockets on the tech’s cart.
  • Socket accessibility: Can a standard socket fit every bolt? A bolt head recessed into a pocket that only a low-profile socket can reach is a shop complaint waiting to happen.
  • Weld nuts vs. tapped holes: Is the fastener accessible from both sides? If a nut has to be held on the back while the bolt is tightened from the front, it’s a two-person job. Use weld nuts, rivet nuts, or tapped holes where possible.
  • Bolt length: Does every bolt protrude at least 2 threads past the nut? A bolt that’s too short doesn’t fully engage. A bolt that’s too long sticks out and catches cables or clothing.
  • Locking: Are critical fasteners locked? Nyloc nuts, spring washers, or thread locker on anything that vibrates. A bolt that works loose in service is a machine that comes back.

6. Pneumatics, Electrical and Cabling

The mechanical designer designs the frame. The electrician and pneumatic engineer hang the components. The design review is where you make sure there’s room for both.

What to Check

  • Valve manifold location: Is the manifold close enough to the cylinders that tubing runs are under 2 meters? Long tubing runs slow the cylinders and waste air.
  • Cable management: Is there a cable tray or drag chain path for every moving cable? Wires draped over the frame don’t last. Cables that flex at a fixed point fatigue and break.
  • Sensor wiring access: Can every sensor’s connector be reached without disassembling something? A sensor with its connector hidden behind a bracket is a maintenance nightmare.
  • Air line routing: Do air lines avoid hot surfaces, sharp edges, and pinch points? A PTFE tube running over a weld bead gets cut. A tube running near a heat source gets soft and leaks.
  • Panel location: Is the electrical panel accessible from the front (or the side the customer specified)? Can the panel door open fully without hitting the frame or a wall?

7. Dimensional and Tolerance Checks

This is the engineer’s domain. But even here, a second pair of eyes catches things.

  • Critical datums: Is the part located on two datums? Is the clamping on a non-locating surface? (Clamping on a datum surface pushes the part off the locator.)
  • Tolerance stack-up: For precision stations, has the tolerance stack-up been calculated? A ±0.1 mm here and ±0.05 mm there adds up to ±0.3 mm by the time the part reaches the next station.
  • Clearance vs. interference: Has the assembly been checked through its full motion envelope? A static assembly doesn’t show what happens when the gripper opens while the feeder is indexing.
  • Gauge pins and dowels: Are dowel pins accessible for removal? A dowel pin you can’t push out destroys the part when you try to repair it. Add a threaded extractor hole.

Running the Review: How to Make It Actually Useful

A design review is a meeting, not a form. Here’s how I run one:

  1. Walk the model in 3D, not 2D drawings. Open the assembly, animate the motion, and trace the part path. People see things in 3D that they miss on drawings.
  2. Bring the shop foreman and the electrician. They see problems the designer doesn’t know exist. The foreman knows what’s hard to weld. The electrician knows what’s hard to wire.
  3. Assign each item an owner. “This needs to be fixed” is not an action. “Zhang to add cable tray to the weld model by Friday” is an action.
  4. Re-review the fixes. The review isn’t done until the open items are closed. A list of action items with no follow-up is a wish list, not a review.

The Bottom Line

A non-standard equipment design checklist isn’t about bureaucracy. It’s about not discovering on the customer’s floor that the operator can’t reach the e-stop, or that the sensor is buried under the gripper. The items on this list are the ones that come back as punch-list items — every single time, on every machine, if you don’t check them.

Bring two people who didn’t draw the machine. Walk the model. Run through the list. The hour you spend in the review saves you a week on the floor. That’s the best return on an hour you’ll get on any project.