12 KUKA Delta Robots Sorting Cookies: Air Circuits, Crumbs, and Cycle Time Challenges

Title

12 KUKA Delta Robots Sorting Cookies: Air Circuits, Crumbs, and Cycle Time Challenges

Meta Description

An engineer built a cookie pick-and-place line with 12 KUKA Delta robots, facing three major challenges: low suction cup pick rates, oil and crumb contamination, and mismatched placement cycle times. This article analyzes the root causes and explores whether larger cylinders and speed adjustments can really solve the problems.

Article Content

An engineer shared his cookie pick-and-place line design on Reddit: 12 KUKA Delta robots are responsible for loading cookies into trays. The line is already in operation, but three thorny issues have surfaced—insufficient suction cup pick rates, oil and crumb contamination, and a mismatch between the robots' placement cycle time and the conveyor speed. The original author admits he is "a bit lost" and is asking the community for help. Although the post is brief, it touches on several very typical engineering pain points in food automation production lines.

Problem 1: Suction Cups Can't Pick Uniformly—Is a Larger Cylinder the Right Answer?

The original post mentions that not all cookies can be successfully picked. He suspects it's related to airflow and plans to make the cylinders in the gripping system larger to let more air through.

This line of thinking is reasonable in direction, but requires more detailed diagnosis. Suction cup pick failures usually have three causes: insufficient vacuum flow, poor contact between the suction cup and the cookie surface, and air leakage due to the porosity of the cookie itself. Cookies are typical breathable materials, especially products with patterns or micropores on the surface. Simply increasing cylinder volume may not improve effective vacuum—if the suction cup lip seal is not tight, a larger cylinder only delays the speed at which vacuum is established.

What is more worthwhile to do first is to measure the actual vacuum level and response time, rather than directly swapping in larger components. In addition, the suction cup material (silicone hardness), lip shape, and acceleration curve at the moment of picking all affect the pick success rate. Community members in similar discussions have pointed out that bellows suction cups commonly used in the food industry adapt better to uneven surfaces, but they need to be matched with sufficient vacuum generator flow.

Vacuum Generators and Effective Vacuum Level: In pneumatic systems, the component that generates vacuum is usually not a "cylinder" but a vacuum generator (venturi type or electric vacuum pump). Venturi vacuum generators use compressed air flowing at high speed through a restrictor to create negative pressure. Their key metrics are suction flow (NL/min) and maximum vacuum (-kPa). Whether a suction cup pick succeeds depends on whether sufficient pressure differential can be established between the suction cup and the workpiece within a very short pick window (usually tens to hundreds of milliseconds). For porous, breathable materials, air will continuously leak from inside the workpiece. At this point the system is in a "dynamic leakage" state, and the actual vacuum level is far below the nameplate value. If increasing cylinder volume means increasing the supply air flow of the vacuum generator, it may help shorten response time; but if it only means increasing the bore of the actuation cylinder, it does not directly help the speed of vacuum establishment. From an engineering perspective, more attention should be paid to the balance between the vacuum generator's suction flow and leakage flow, as well as the matching of suction cup volume and piping volume—the longer the piping and the larger the volume, the slower the vacuum is established. It is recommended to use a vacuum pressure switch with a digital display to record the pressure curve at the moment of picking, so as to determine whether the issue is insufficient flow or excessive leakage.

Problem 2: Oil and Crumbs—the Hidden Cost of the Suction Cup Approach

The second problem is more troublesome: when picking cookies, oil and crumbs enter the suction cup system. The original author tried a mechanical gripper similar to a "claw" to avoid contamination, but the cookies are arranged tightly on the conveyor, and when the gripper picks one, it easily crushes adjacent cookies.

This is a classic dilemma in food pick-and-place: vacuum suction cups are clean and efficient but vulnerable to contamination; mechanical grippers are not picky about surfaces but require larger operating clearance. The difficulty in the original post is that product spacing is too small, which naturally puts the gripper approach at a disadvantage. Possible compromise paths include: installing replaceable filter cotton or micro-filters at the front end of the suction cup to intercept oil contamination outside the vacuum circuit; or adopting a combination of "suction cup + slight blow-off," using positive pressure blow-off at the moment of release to help the cookie detach and clean the suction cup surface. However, these solutions all need to be validated for their impact on cookie integrity.

From a line maintenance perspective, oil entering the vacuum generator will shorten its life, and crumbs may clog the air circuit. If the contamination problem is not solved, increasing cylinder size will instead amplify the maintenance burden.

Filter Selection in the Vacuum Circuit: In food pick-and-place lines, oil and crumbs entering the vacuum circuit are common sources of failure. The standard practice is to install a micro vacuum filter between the suction cup and the vacuum generator (such as sintered metal or replaceable cartridge type), with filtration accuracy usually selected at 5–40 microns. However, a mixture of cookie crumbs and grease easily forms a paste-like clog on the filter surface, causing flow reduction and slower response. Therefore, the filter needs to be designed with a quick-release structure and included in a regular maintenance plan (for example, inspection every shift and replacement every week). Another approach is to use high-flow blow-off backflushing: briefly apply positive pressure into the suction cup after release to blow crumbs away from the suction cup lip and piping inlet, reducing contaminants entering the circuit. Note that the blow-off pressure and time must be controlled within a range that does not affect cookie positioning and integrity. In addition, oil contamination will accelerate wear and clogging of the venturi inside the vacuum generator and shorten its life, so the filtration solution not only protects the suction cup but also protects the entire vacuum generation system.

Problem 3: Cycle Time Mismatch—Is Adjusting Speed Really Enough?

The third problem is that the robot "waits too long before placing cookies." The technicians' solution is to simultaneously increase the speed of the tray conveyor and the robot. The original author is skeptical and believes it will not solve the problem.

This skepticism is justified. Pick-and-place cycle time is determined by a series of linked stages: the robot's pick travel, place travel, vacuum establishment and release time, and the product supply rhythm on the conveyor. If the bottleneck is vacuum release—that is, the cookie sticks to the suction cup and does not drop for a long time—then increasing conveyor speed will only make the tray leave faster and actually worsen the waiting. The correct approach is to first locate the bottleneck stage: is the robot motion path too long, is the placement signal triggered too late, or is vacuum break not fast enough?

With 12 Delta robots working together, task allocation and synchronization are also involved. If the pick-and-place task scheduling among robots is unreasonable, individual robots may wait idle while the overall cycle time is dragged down by the slowest one. This kind of problem is difficult to solve by "adjusting both speeds together"; it requires looking at the line's task allocation logic.

Delta Robot Cycle Time and Task Scheduling: The cycle time of Delta robots (parallel robots) is usually measured in "picks per minute," and under standard test conditions (such as a 25/305/25 mm path) can reach 100–200 picks per minute. However, actual line cycle time is jointly constrained by end effector action time (vacuum establishment/break), vision positioning time, conveyor tracking synchronization, and multi-robot task allocation. When 12 robots work together, a common architecture is for a central controller (such as a PLC or robot scheduling software) to dynamically allocate product positions on the conveyor to each robot. The allocation algorithm needs to consider each robot's current load, pick-and-place travel, and available time window. If the allocation logic is simple (such as fixed zoning), some robots may be idle while others are overloaded, and the overall cycle time is dragged down by the slowest station. From an engineering perspective, cycle time simulation (such as RobotStudio, KUKA.Sim) can be used to test different scheduling strategies in a virtual environment, and encoder-synchronized conveyor tracking can be added to the line to ensure that the placement signal aligns precisely with the tray position, rather than relying on simultaneously increasing two speeds and "hoping for the best."

Common Judgment from Multiple Perspectives

Although this post comes from a single source, similar problems recur in Reddit discussions about food pick-and-place lines. Community consensus tends to be: for vacuum picking of cookie-type products, success rate and contamination control are often more critical than robot body selection; and cycle time problems are mostly not about speed parameters but about signal timing and vacuum response. The original author's two intuitions—"increase cylinder size" and "adjust speed"—both point to symptoms rather than root causes.

For engineers designing similar lines, the value of this post lies in reminding them: Delta robots themselves are mature products, and what really takes skill is matching the end effector to the process. Suction cup maintenance access, filtration solutions, and per-station cycle time margin should be considered at the drawing stage, rather than remedied after the line is running.

One-line Summary

A 12 KUKA Delta robot cookie pick-and-place line exposed three major engineering pain points: low suction cup pick rate, oil contamination, and cycle time mismatch. The root causes are mostly in end effector and process matching rather than the robot body.

Paragraph Summary

An engineer shared on Reddit his design for a 12 KUKA Delta robot cookie pick-and-place line. The line is in operation but has exposed three typical problems: insufficient suction cup pick rate, oil and crumb contamination of the vacuum circuit, and a mismatch between the robots' placement cycle time and conveyor speed. The original author plans to increase cylinder size and simultaneously increase conveyor and robot speed, but the article argues that both intuitions point to symptoms rather than root causes. Cookies are porous, breathable materials, so pick failures should first be diagnosed by measuring actual vacuum level and response time; oil contamination requires replaceable filters at the front end of the vacuum circuit and reserved maintenance access; cycle time bottlenecks may lie in vacuum release or task allocation logic, and blindly increasing speed may worsen waiting. The article emphasizes that Delta robots themselves are mature products, and what really takes skill is matching the end effector to the process. Suction cup maintenance, filtration solutions, and per-station cycle time margin should be considered at the design stage.