If you design vision inspection machines, you've probably faced this puzzle: the conveying axis runs over a meter, the lifting axis moves just a few hundred millimeters, and each axis has different speed, load, and precision demands. Which linear module goes where — without overspending or underspecifying?

We recently supplied a complete linear motion package for a Chinese manufacturer of preform inspection machines (machines rated for 48,000 parts/hour). Here is the full breakdown of that belt-driven + rack-and-pinion combo: why long travel gets a belt module, why the vertical axis gets a rack — and the three mistakes that cost machine builders the most rework time.

Layout of a preform vision inspection line (AI-generated illustration, not an actual project photo)
Layout of a preform vision inspection line (AI-generated illustration, not an actual project photo)

The Two Axes That Matter

The motion section of this machine is simple — two axes:

Axis Function Solution Stroke Key components
X (horizontal transfer) Move preforms into the inspection station Belt-driven module 1200mm 8M AT-type timing belt + planetary gearbox + 6 sliders
Z (vertical lift) Raise/lower camera & gripper Rack-and-pinion module 300mm Helical rack drive + 90° bevel gearbox + 4 sliders

Why the Long Axis Gets a Belt-Driven Module

Strip a belt-driven linear module down and you find four things: a motor + gearbox driving a pulley, a timing belt that moves the carriage, guide rails for direction, and photoelectric sensors for positioning. No ball screw, no gear mesh — which makes it the natural choice for "fast and far":

  • Flexible stroke. The 1200mm X-axis here is built from two 600mm base sections joined with dowel-pinned splices. Single-piece bases that long are expensive to ship and often won't fit through a standard elevator; a properly pinned splice gives up nothing in accuracy.
  • Speed. Belt modules routinely run 1–2 m/s — several times faster than ball-screw units (typically under 0.5 m/s). High-throughput inspection lines can't live without it.
  • Lower cost at long stroke. At equal stroke, belt systems run 20–40% cheaper than ball screws, with no critical-speed or screw-droop issues to engineer around.

Note the six sliders on the X-axis — three times the usual count. The gripper and product weight all ride on that carriage; more sliders share the load, which means less deflection and longer rail life.

Belt-driven module (horizontal) + rack-and-pinion module (vertical lift) combo structure (AI-generated illustration)
Belt-driven module (horizontal) + rack-and-pinion module (vertical lift) combo structure (AI-generated illustration)

Why the Lifting Axis Gets a Rack-and-Pinion Module

For a 300mm vertical stroke, why not a belt — or a ball screw?

  • No ratcheting under gravity. Belt drives rely on tooth engagement; under constant vertical load they carry a skip-tooth risk. A rack-and-pinion is a hard mesh, and paired with a brake motor it holds position even on power loss.
  • Accuracy where it counts. Helical gears mesh with high contact ratio and minimal backlash, delivering sub-millimeter repeatability for station positioning — with far less long-term wear than straight-cut spur gears.
  • The 90° bevel gearbox is the quiet win. Instead of a motor sticking straight up from the top of the module, a right-angle planetary gearbox lays the motor horizontally. On an inspection machine, that saved height is where the camera, lighting, and service access all live.

Selection Table (Steal This)

Condition Recommended solution Typical application
Stroke >800mm, horizontal, speed priority Belt-driven module Conveying, loading/unloading, long transfers
Stroke <500mm, vertical, holding required Rack-and-pinion + brake motor Lift axes, Z-axis positioning
Stroke <1m, accuracy ≤±0.02mm Ball screw module Precision dispensing, fastening
Stroke >2m or heavy loads Rack-and-pinion + oversized rails Gantry systems, heavy frames
Vertical axis with tight headroom Rack-and-pinion + 90° gearbox Inspection machines, filling machines

Three Mistakes That Cost Real Money

  1. Splices without dowel pins. Bolt-only splices creep. Six months in, the joint shifts a few hundredths of a millimeter, sliders clunk across the seam, and your camera images blur. Pinned + ground splices are non-negotiable.
  2. Photoelectric limit switches with no mechanical backup. Dust kills optical sensors. When one fails mid-cycle, the carriage slams into the end stop. Every axis in this build carries both external photo sensors and mechanical stop blocks — tens of dollars of insurance against a week of crash repairs.
  3. Specifying a belt on a vertical axis to save budget — and watching the camera free-fall on power loss. Holding requirements must be settled at selection time: rack-and-pinion with brake motor, or belt with a brake. Never bare.

Bottom Line

Specifying linear motion for a vision inspection machine comes down to one sentence: let the belt module own horizontal speed, let the rack-and-pinion own vertical holding. This combo is proven at 48,000-parts-per-hour throughput and transfers directly to cap inspection, part feeding, rejection, and palletizing cells.

Specifying a machine of your own? Download our product catalog for full specifications on belt-driven modules, rack-and-pinion modules, and linear guides — or contact us with your stroke, load, and cycle time, and we'll send back a configured proposal.

Further reading: - Belt vs. Ball Screw: How to Choose a Linear Module - Rack-and-Pinion Modules for Long-Stroke Applications