Enhanced Automation Drives More Capable Robotics in Injection Molding

During the 2018 NPE, visitors observed a shift from isolated molding machines to integrated automated cells that combine multiple upstream and downstream functions. According to Chris Parrillo, national sales manager at Yushin America, "We’re seeing more automation beside the press—enhanced inspection, packaging, assembly, and product serialization through hot stamping, laser engraving, barcode labeling, and RFID."

Dino Caparco, Yushin’s engineering operations manager, notes, "Takeout robots are increasingly employed, enabling higher speeds and more sophisticated end‑of‑arm tooling for tasks such as insert molding, in‑mold labeling, and part removal."

Consequently, machine vision has become increasingly integral to robotic systems. Vendors cite molders’ heightened quality focus, coupled with falling costs and miniaturization of vision hardware, as the driving factors. Cameras mounted on molds, frames, or robots now routinely verify short shots, color accuracy, degating, insert loading, and full cavity ejection. Vision systems are also being deployed for 100% part inspection, measuring dimensions, label alignment, and surface defects, with up to 16–18 parameters evaluated per part.

While 2015’s NPE showcased vision‑guided robots performing playful tasks like basketball or miniature golf, the real‑world application lies in precise part location, orientation, stacking, and packing on conveyors. Jim Healy, VP of sales and marketing at Sepro America, observes that current vision solutions are largely custom, but the trend points toward more plug‑and‑play systems.

Healy attributes the dominance of six‑axis articulated robots at this year’s NPE to the broader shift toward cell automation. Although highly versatile, jointed‑arm robots are generally preferred for downstream tasks rather than mold‑pulling. Compared to linear robots, they typically offer lower speed, reach, payload, and require larger guarded work envelopes, which increase cost and programming complexity. However, they can excel in low‑headroom scenarios.

Nevertheless, vendors such as Sepro, Yushin, and Wittmann Battenfeld contend that linear robots equipped with a servo wrist can achieve five to seven degrees of freedom, rivaling jointed‑arm systems. At NPE, linear robots were observed functioning solo, in tandem (passing parts between units), or in collaboration with jointed‑arm robots for downstream operations.

Downstream cell tasks also benefit from specialized robots that may be new to many attendees. SCARA robots—compact, high‑speed machines—excel at pick‑and‑place of small, lightweight parts, offering three or four axes and up to 200 picks per minute while accessing tight spaces. Another category is the spider, delta, or parallel robot, featuring three arms linked to universal joints at the base. With three to six degrees of freedom, they can achieve up to 300 picks per minute. These rapid systems can reposition multi‑cavity parts immediately after demolding by a linear robot, eliminating the need for a buffer zone within the press cycle.

A newer entrant, the collaborative robot—or cobot—has attracted significant interest. Cobots are characterized by relatively low cost, intuitive lead‑through teach‑mode programming, and, crucially, safe operation alongside humans without the need for hard guarding in certain contexts. Typically built on six‑ or seven‑axis jointed‑arm architectures, they employ integrated vision and torque sensing to detect accidental contact and immediately stop movement. While ideal for downstream pick‑and‑place tasks rather than machine tending, many vendors now offer cobots, and established Cartesian robot manufacturers view them as a valuable portfolio extension.