How robotic developments have improved repeatability, reproducibility and throughput in small part measurement

by

Heidi Quinlivan, sales and marketing vice president, New Scale Robotics

More than 62 years ago, the first industrial robot, known as a fixed robot, initiated a wave of manufacturing automation. Just 13 years ago, the first commercial collaborative robot (cobot) brought the benefits of industrial robotics to high-mix, small-batch manufacturing. As automation technology has advanced, so has the complexity of problems to be solved. Yet some of the simplest challenges continue to confound and frustrate manufacturers. Among these is automating measurement of small parts for quality assurance (QA) and quality control (QC). 

This article explores the evolution of QC measurement techniques, including recent end-of-arm tooling developments that allow cobots to perform QC measurements. Cobots, like fixed robots, automate, dull and repetitive work, keep humans safe from injury and deliver more consistent performance. They also improve measurement repeatability and reproducibility. In sum, cobots offer manufacturers a competitive advantage in the race to maintain high quality and reduce costs.

From fixed robots to cobots, and from production to QC

The first fixed robot launched in 19591. Since then, it has become commonplace in almost every manufacturing industry. Fixed robots afford efficiency and cost-saving benefits but have significant drawbacks, for example:

  • their large footprint and safety cages mean that their installation often involves expensive renovations and they are not easily relocated;
  • their operation alongside humans necessitates extensive safety protocols; 
  • they cannot be easily reprogrammed for different tasks;
  • their repair and software upgrades can halt production for days;
  • they have a typical return on investment (ROI) of 8–10 years, and advances in technology can render them obsolete before ROI has been achieved; and
  • their operation, repair and programming require technical expertise and advanced education, meaning there are few opportunities for workers to retrain or advance as their jobs are changed by automation.

The cobot made its debut in manufacturing in 2008. Cobots mitigate many of the drawbacks of fixed robots, namely:

  • they occupy a much smaller footprint and require little to no specialised safety equipment, so they can be easily moved from one location to another;
  • they are designed to work in tandem with humans;
  • they have a typical ROI of less than a year; and
  • they can be programmed and reprogrammed quickly using a tablet, laptop or teach pendant, meaning any worker familiar with a smartphone can easily learn to program them and there are more opportunities for up-training and advancement of existing employees. 

Two key features of cobots are (1) the availability of standard, plug-and-play end-of-arm tooling for different tasks, and (2) the ease with which they can be taught to perform these tasks.

Numerous vendors have developed end-of-arm tooling for cobots manufactured by Universal Robots (UR). Tools include electric grippers, pneumatic grippers, screwdrivers and tools for cutting, sanding, welding and more. Now, a new type of tool known as a robotic precision caliper extends the applicability of UR cobots to the tasks of sorting and measuring small parts for quality control.

Evolution of small part measurement in QC

Prior to the Industrial Revolution, factory owners made their own machines for manufacturing. They could make custom replacement parts for their machines as needed. By the 18th century, increasing mechanisation required more uniform replacement parts and standardised measurement became the norm2.

As mass production became more prevalent, so did the need to produce parts with consistent dimensions, within established tolerances. This led to widespread use of handheld measuring tools for QC/QA in manufacturing.

Calipers and micrometers are the most common handheld measuring tools. Operator skill finesse is required to position the measuring tool and part together and interpret the results.

Using handheld calipers for QC inspection of a large number of parts is tedious at best and can present a risk of repetitive strain injury (RSI). Furthermore, handheld digital caliper measurements vary both among operators, because of different skill levels, experience and techniques, and over time for each operator, due to skills development, fatigue, distraction, health, mood and other human factors. 

Manual data entry of measurement results is a significant source of errors. A recent survey of 260 manufacturers, including some of the world’s largest, revealed that 75 percent are still collecting data manually. Of these, an astounding 47 percent still rely on pencil and paper3. Handheld smart gauges can be used to transmit data electronically to a computer, helping to reduce errors and provide better records for statistical process analysis4, but the problem of human variability in the measurement technique remains. 

Laser scanning and machine vision

The 1970s saw breakthroughs in sensors, machine learning, and laser scanning and machine vision systems. These advances were applied to meet the demand for higher precision and tighter tolerances on parts. For the first time, a amachine rather than a human was responsible for accuracy and measurement.

Laser scanning and machine vision measurements depend on the reflective and transmissive properties of the part being measured and the properties of the lighting source. Shiny surfaces and small surface variations can create false errors in acceptable parts or mask real defects. Laser scanning systems have embedded light sources but require consistent surface reflectivity, therefore they are unable to measure small features if their line-of-sight (LOS) is occluded.

Furthermore, machine vision systems require an experienced person to optimise and validate each measurement. Often, only computer technicians or engineers have the requisite expertise, meaning that opportunities for advancement from the shop floor are limited for those that do not have post-secondary education in the right areas.

However, the advantages of optical measurements outweigh their limitations, and the applications for them are growing.  Non-contact optical measurement is a good choice for small, thin and fragile parts that cannot be measured through contact optical measurement. It is also useful for fast in-line inspection. Laser scanning allows for fast 3D measurements of large, complex shaped parts with micrometre precision; and machine vision is ideal for 2D measurements of backlit thin parts with micrometre precision.

Coordinate measuring machines (CMMs)

The first coordinate measuring machine (CMM) was a two-axis device developed by the Ferranti Company in the 1950s. The first three-axis CMMs appeared in the early 1960s. They consisted of 3D tracing devices with a simple digital readout displaying the X, Y, Z position. Today’s CMMs evolved from these.

CMMs measure the geometric properties of an item to precise specifications and can be used either manually or via a computer program5. They can generate real-time reports of deviations and risk tolerance fails.

CMMs are expensive instruments. They require heavy and stiff structures to achieve their fundamental precision, so they are not easily relocated once installed. Furthermore, they are susceptible to measurement errors due to moisture and temperature variations and vibration, so must be located in environmentally controlled rooms.

Other disadvantages of CMMs include: the significant training needed to program each part and collect measurements; the frequent computer and software upgrades and re-training needed to remain compatible with the newest CAD/CAM programs and avoid obsolescence; the unique fixturing needed for each part; the need for insertion and removal of each part by a person; and the typical several minute cycle time for each part, as this is not a fit for high-volume manufacturing.

However, CMMs are the best choice for critical and complex 3D measurements of parts with micrometre precision. In high-volume manufacturing, they make detailed first article inspections (FAIs) that validate a process and certify golden parts that may be used for faster in-line comparative measurements. 

Cobots for QC        

Just as cobots brought the benefits of robotic automation to small-batch, high-mix manufacturing, they are now bringing those same benefits to the QC lab. Three recent developments have made this possible, namely:

  • robotic precision calipers;
  • smaller, more lightweight grippers; and
  • a system that allows multiple calipers and grippers to be mounted on a single cobot and controlled by a unified program.

These developments support exciting new applications for cobots in the QC laboratory.

QC cobot application 1—automated small part measurement

New Scale Robotics (NSR) has brought the aforementioned three developments together and paired them with a UR3e cobot arm to create the Q-Span system for automated small part measurement.

The Q-Span system can automatically pick parts, measure multiple dimensions on each and make real-time decisions such as pass/fail based on the results. It can then sort the parts into output trays and send all the data to a PC for later analysis.

A typical Q-Span system has one robotic gripper for part handling and one or two robotic precision calipers to take measurements. In many cases, one tool serves as both gripper and caliper. Each caliper is equipped with metrology fingers specific to the part and the dimension to be measured. Dimensions that can be measured include length, width, thickness, outer diameter and inner diameter.

The important features of the gripper and caliper/s are their small size, their light weight and the high precision they enable of the fingertip motion with position feedback to the robot. Their small size and light weight mean that three devices can be mounted on UR’s smallest cobot. This allows the system to occupy minimal lab space and perform multiple processes with fewer large moves. The result is high throughput and safety for human operators.

The gripper and caliper of a Q-Span system replace a manual digital caliper and afford around twice the precision. When using the appropriate fingertips and established metrology practices, part measurement resolution is 0.0001 in. (2.5 μm). The +/-3 sigma repeatability and accuracy are factory-certified to be less than +/-0.0002 in. (5 μm) and +/-0.0006 in. (15 μm) respectively. 

Once the Q-Span system is taught the proper force, position and orientations needed to make precise measurements, it can repeat the procedures indefinitely without fatigue or distraction and can continue to operate in off-hours. It can be taught procedures for many different parts, and then changeover from one part to another can be accomplished in under an hour.

Q-Span training and operation are straightforward and do not require advanced technical skills or education. This provides new opportunities for existing QC technicians to advance and receive up-training. The chance to do higher-level work with less tedium helps companies retain valuable employees.

A Q-Span system has an average ROI of less than 11 months based on labour savings alone. The ROI is faster when considering the ability to add QC capacity without increasing headcount.

QC cobot application 2—tending of CMMs

A cobot can be used for the tending of CMMs. OptiPro Systems, a manufacturer of grinding, polishing and metrology machines for precision optics, has demonstrated using a Q-Span system to automatically load and unload a CMM. 

This Q-Span system is configured with one precision caliper that both picks and measures the part. The caliper measurements are used to increase throughput and utilisation of the CMM. If the preliminary diameter measurement is within range, the Q-Span system loads the part into the CMM. If it is out of range, the Q-Span system can place the part directly into the fail output tray, thus avoiding wasted CMM cycles.

After a part is loaded, the CMM uses the Q-Span system’s measurement to determine which routine to perform. It then runs the full routine and sends its data back to the Q-Span system, which unloads the part and places it into the appropriate output tray, pass or fail.

Robotic tools replace manual digital calipers and other gauges

Real-world metrology must measure parts that are complex and it is rare that digital caliper measurements are sufficient for all features. Q-Span workstations can be expanded to include complementary measuring instruments and tools with greater precision. 

A Q-Span workstation can incorporate other gauges, with digital output, that measure a variety of part features including depth, height, inside diameter and outside diameter. Examples of complementary instruments and tools include, laser height gauges, 2D laser scanners, air gauges, laser height gauges, linear variable differential transformer (LVDT) height gauges and mechanical snap gages.

Benefits of cobots for automation in QC

Using robots to automate measurements eliminates human variation and reduces delays, scrapped material and cost. Using cobot system architecture ensures maximum flexibility for high-mix, small-batch manufacturing. The result is greater inspection capacity and throughput while increasing employee engagement, skills and retention.

Conclusion

Industrial robots have become mainstream automation components in the manufacturing sector, performing a multitude of repetitive tasks that require accuracy, precision and consistency. Cobots afford a number of advantages over traditional fixed robots.

Fixed robots have large footprints, their installation often involves expensive renovations and they are not easily moved. Their operation, repair and programming are highly specialised and can only be undertaken by sufficiently qualified persons. Cobots, on the other hand, have a smaller footprint, and they are easily moved from one location to another. They can be easily programmed by any worker who is familiar with a smartphone, providing opportunities for up-training and advancement of existing employees.

Moreover, and perhaps most importantly, recent end-of-arm tooling developments, such as robotic precision calipers and smaller, more lightweight grippers, are allowing the advantages of cobots to be realised for small part measurement in QC labs.

New Scale Robotics

www.newscalerobotics.com

1Joseph Engelberger, the father of robotics [web page]. A tribute to Joseph Engelberger. Robotics Industries Association.

Available at: http://bit.ly/3ckBEMD

2The history of measurements [article]. MTI Instruments.Available at: http://bit.ly/2OE7Mmc

3InfinityQS (2020). Manual data collection delays digital transformation for manufacturers [article/sponsored content]. April 13. Quality Magazine.Available at: http://bit.ly/3ch0Y6k

4Schuetz, G. (2019). Choosing the right smart handheld gage [article]. March 1. Quality Magazine.Available at: http://bit.ly/3l2HHJL

5Brown, D. (2016). A history of metrology—a look at measurement solutions and metrology devices from Galileo to today’s optical systems [article]. October 3. Design Engineering.Available at: http://bit.ly/30yu5wc