Direct Part Mark Identification Issues
Data Matrix code, a two-dimensional symbology, permits manufacturers to encode text and numbers on a part
By Justin Testa, Senior Vice President of ID Products
Many medical-device manufacturers ensure complete quality control over their production processes by directly marking the surfaces of parts and products with machine-readable codes, reading the codes, and tracking the items as long as they are useful. This process is known as direct part mark (DPM) identification. Not only does marking medical devices help to improve production and inventory control at the manufacturer, the FDA has recently noted that a compelling patient safety interest also lies in requiring codes on medical devices that could be subject to recalls.
Programs that employ DPM ID for traceability encourage cost effectiveness and supply-chain efficiency, and hold promise for reducing preventable medical errors, speeding recalls, improving product containment, and fostering clinical product and service innovations. Fast, accurate, reliable DPM identification is a challenge, however, because codes can be difficult to read due to low contrast, variations in part surfaces, partial damage caused by process, handling, and environmental conditions.
Because the quality of a direct-marked code is so critical to the success of lifetime part traceability, many manufacturers view the code as a critical attribute of the part itself. A loss of the part’s identity due to poor mark quality means that the part cannot be processed or used in the supply chain. And, when a code is unreadable, an automated machine may stop working. Consequently, ID readers must provide accurate, reliable performance, and high read rates to prevent machine interruptions, while reading fast enough to keep pace with high-throughput equipment.
Advances in the power of digital-signal processors, imaging sensors, and decoding algorithms have resulted in the production of cost-effective ID readers that deliver the reading results required to achieve the objectives of medical device manufacturers. Implementing a successful DPM application compels manufacturing engineers to consider several factors.
Code selection can be critical. Data Matrix is a public domain symbology that is defined by ISO’s 16022 symbology specification. It’s the most widely supported symbology for DPM applications involving metal, glass, ceramic, or plastics. Just to be clear, Data Matrix is not some code that our company prefers. We support an expanding list of 1-D barcodes and 2-D codes, among them Code 128, Code 39, UPC, EAN, 12 of 5, and QR Code. But Data Matrix is the recommended standard for DPM on medical devices, and for DPM in most other industries as well.
The newest and most standard version of Data Matrix error correction, ECC200, supports advanced encoding and error checking with error-correction algorithms, which allow decoding even if up to 60% of the mark is damaged. Marking and reading equipment suppliers have invested significant R&D resources to improve the performance of ECC200 supporting equipment. Although Data Matrix supports a number of different formats and error-correction methods that include ECC000, 050, 080, 100, 140, and 200, all industry standards and guidelines for DPM applications are based on the ECC200 format of Data Matrix.
The actual Data Matrix code is a two-dimensional symbology that enables the encoding of text, numbers, files, and actual data bytes. It has several advantages for DPM applications, including small size, high data-encoding capacity, and error correction. Data encoding refers to the amount of information stored in the matrix when Data Matrix code is generated. There are 24 square formats and six rectangular formats available in ECC200 to provide users the flexibility to encode between six and 3116 digits in a single code.
Code size can affect readability. It’s generally determined by the amount of data to be encoded, module (cell) size, and surface roughness of the area on the part where the code will be applied. When trying to comply with an industry specification, an application specification will define the content of the information that must be encoded, and the size of the code needed to be in compliance. The decision as to what information to encode is typically based upon the end user’s requirements for the traceability project.
It may sound obvious, but it’s important to consider the space that is available on the part. Limited space may require using a small Data Matrix code as a “license plate” just to identify the part. In this case, a centralized database containing manufacturing and historical data referring to the item is updated as it’s identified during manufacturing and supply chain processes. When space isn’t an issue, take advantage of the code’s large data capacity. Place enough information on the part to make the code a true portable database.
The choice of marking method is typically incorporated into the component design. Deviations from this design may require engineering change approval. The primary methods used to produce machine-readable symbols for DPM include dot peening, laser marking, electrochemical etching, and ink-jet printing. Select a marking method by considering part life expectancy, material composition, environmental wear and tear, and production volume. Other considerations include surface texture, the amount of data to be encoded on each part, and the available space and location of the mark on the part.
Give some thought as to where to place a mark. The code’s location on the part can directly affect its readability. The location should be clearly visible throughout the manufacturing process, and it’s best to mark on a flat site on the part. Also, choose a location where the mark is in a prominent position so that the code reading system can easily view it. Avoid locations where there may be a surrounding surface relief that might affect the illumination of the code by the code reader’s illumination source.
Wherever possible, provide a clear zone around the mark where no features, part edges, noise, or other interference comes into contact with the code. When the mark must be placed on a cylindrical part, take care in selecting the size of the code. Surface curvature can create code distortions and make proper code illumination very difficult. Mitigating the problem calls for a code that takes up no more than 16% of the part diameter or 5% of its circumference.
Readability is a term used to define how easy or difficult it is for a reader to successfully read a code. Because DPM is rapidly becoming a required part of manufacturing, if a code is not readable the part may not be processed, or production equipment may stop. Until recently, manufacturers implementing DPM have tolerated varying levels of read rates, in some cases approaching the upper 90% level. This level of performance, however, is no longer acceptable. It is now imperative to achieve 100% read rates.
Each Data Matrix symbol includes a quiet zone, the finder pattern, the clocking pattern, and the data region. Each individual element is called a module or a cell. The actual appearance of the code depends on the type of mark placed. For example, a DPM code formed with a laser-marking machine or printer would appear with a “continuous L pattern” and square modules; dot peen and ink-jet markers produce codes that have a “non-continuous L pattern,” with a data pattern made up of round modules.
obust and reliable code reading requires a pattern with individual modules consistent in shape and size with each other, and distinctively different in shape and size from other features on the part’s surface. Meeting this requirement can be challenging in DPM applications, however, due to variations in surface texture, variations in part presentation during marking, and variability of the marking machine.
Verification of marking quality is a challenge. With so many manufacturers today implementing DPM identification programs for part traceability, the need to control the part-marking process is becoming increasingly important. The original quality of a DPM code—which serves as the item’s permanent identity—can greatly affect its readability as it travels throughout the manufacturing and supply chain, and ultimately through to its end use.
A verification system can immediately detect a mark-process problem, which could result from poor part fixturing, incorrect settings during part changeover, or damage to the machine, such as a broken tip on a dot-peen machine. Additionally, a code-verification system can also provide feedback during marking that can be used in preventive maintenance. For example, the verifier can monitor tip wear on a dot-peen machine by monitoring dot size, and alert the operator when a pin should be changed.
The most frequently encountered mark reading systems are fixed-mount, presentation, and hand-held. Fixed-mount readers identify parts handled and moved by conveyor, indexer, or robot. Typically, fully automated manufacturing equipment—of the sort used in electronics and automotive facilities—use fixed-mount readers. This type of reader is fixed in a location where the mark can repeatedly be placed in front of the reader in either continuous or indexed motion. A trigger signals that the part is ready for reading. The trigger comes from an external sensor that detects the presence of the part, or by an encoder that knows the part position.
Although a presentation reader is also fixed, it operates in a continuous reading cycle, automatically decoding the mark on the part once the operator places the part in front of it. Presentation readers can quickly read part codes when parts are manually handled. A presentation reader can be implemented with either a fixed-mount reader or hand-held reader.
Hand-held readers are preferred where part handling is not automated, or where parts vary greatly in size. They are used in job shops, QC test stations, and in logistics areas. Hand-held readers come in either tethered (corded), or cordless configurations. Cordless operation may be required where part size or position impose a practical limitation to cord length.
Manufacturing engineers should select readers that tolerate a wide range of normal code variability, because consistent code reading is critical in DPM ID applications. Distortions to the code are common, and result from part material composition, variations in part presentation, or variability caused in manufacturing. A set of sample parts representative of the range of mark quality that a reader will need to handle should serve as the basis for a preliminary test of a reader’s rate. A more extensive pilot test is advisable, so more read-rate statistics can be gathered and analyzed.
The reader should also return a result quickly. Although the function of an automatic reader is to eliminate data errors, the implementation of the reader cannot slow down the process. Until recently, hand-held readers were very slow when used to decode direct-marked parts, often resulting in a no-read signal. To eliminate operator frustration, it’s important to look for hand-held readers that provide the “laser-like” scanning performance that we are all accustomed to in retail applications across the entire set of representative sample parts provided for a test. A “trigger-to-good-read beep” should be consistently heard in less then one second on all parts to gain operator acceptance on the factory floor.
In fixed-mount DPM applications, the DPM reader should offer both serial and network communications, because results are usually sent to process equipment or a database over the factory network. Serial communications are typically used in applications where the read or verification results stay local to the workcell or factory-automation equipment. Network connectivity lets the reader communicate decoded results to PCs and databases at the enterprise level.
Finally, as more and more ID readers are used throughout manufacturing, it becomes important to have a centralized way of managing them. Make sure the ID reader will allow you to manage and control activity over the network from remote locations in the plant and beyond.
Connectivity methods used with hand-held readers depend on whether the reader is tethered or cordless. Tethered readers send results through a keyboard wedge interface. It emulates keyboard keystrokes. This makes integration to a PC simple. Alternatively, communications can be made over an RS232-C interface. A cordless hand-held reader uses wireless technology, such as Bluetooth, to communicate to the base PC station or other controller. For more information on Cognex part mark reading systems.
This article was first published in the May 2006 edition of Manufacturing Engineering magazine.