In the Sept 12, 2003 issue of the Data Capture Report, Intermec took issue with some of the statements made by Symbol Technologies concerning the relative merits of laser and linear imaging technology for reading barcodes. The statements made by Symbol were in response to a comparison between the Symbol LS-4000 laser scanner and the HHP IT 5600 linear imager, which was published in an earlier issue. In that earlier issue HHP claimed that, with the introduction of the IT 5600, linear imagers had finally achieved performance that is better than laser scanners because the 5600 was able to boast working range that is better than the popular LS-4000.
It is Symbol's position that scanning performance cannot be equated to working range or to any other single performance parameter. Working range is one parameter that contributes to performance, but it is not performance all by itself. The real measure of scanning performance is productivity. In other words, the scanner that provides the highest performance in any application, is the scanner that allows its operator to scan the largest number of barcodes in a given period of time.
Lack of enough working range for a given application can subtract from the productivity that can be achieved with a scanner, but increasing range beyond that which is needed does not improve productivity. Therefore, increased range does not equate to increased performance. We stated that many years of experience has shown us that the working range of the LS-4000 is already more than adequate for the markets into which it is intended to sell, so a scanner with more range doesn't automatically become better, especially if it isn't as good in some other ways.
To further support our claim that working range and performance are not synonymous, we discussed several other performance attributes, all of which need to be optimized in a single scanner if that scanner is to achieve maximum productivity. The purpose of discussing these other attributes was to demonstrate that a scanner that maximizes one attribute, including an obvious one such as range, will not necessarily provide the best productivity unless the other less obvious attributes are also optimized. In total, our remarks explained why a high performance laser scanner, such as the LS-4000 will still provide the best overall productivity in most scanning applications, even though the IT 5600 has more range.
In the Sept. 12 issue, Intermec disagreed with several of the statements we made. Below are our responses to the specific objections and issues raised by Intermec.
Motion Sensitivity
Symbol claimed that it is important that a scanner be able to read well even if it is in motion with respect to the barcode while scanning. Intermec disagrees, saying that they believe that users of hand held scanners always hold the scanner stationary during scanning. While it is true that most users don't intentionally move the scanner, some untrained users do wave the scanner up and down or left and right while scanning, apparently because they think that is how a scanner is supposed to be used. It is not possible to predict how an untrained operator will use a scanner, so it is best to use one that will work well even if used in an unexpected way. This, however, was not what we were talking about when we discussed motion sensitivity.
The real issue stems from how users behave when they are in a hurry and need to scan a lot of barcodes in a short period of time. Users in busy scanning environments learn that they can achieve the greatest productivity by rapidly moving the scanner from one barcoded item to the next, scanning as they go. If they don't need to come to a complete stop in front of each barcode the scanning process will be faster and less tiring.
If a linear imager is being used, even a small horizontal motion (in the direction perpendicular to the printed bars), of the kind that occurs while the scanner is being moved from one barcode to the next, can make the scanner hesitate. This makes the scanner feel sluggish and generally reduces the overall productivity that can be achieved. This sensitivity to motion is why a linear imager might feel very snappy in a carefully controlled lab environment, but will sometimes feel sluggish when a user in a real environment is in a hurry. The faster the user tries to go, the less careful he will be to completely stop the scanner in front of each barcode, and the more frequently the linear imager will hesitate. A laser scanner, on the other hand, does not lose snappiness when the user is in a hurry.
Scanning Speed
Intermec points out that linear imagers have faster scan rates than most hand held lasers. Specifically they are speaking about the number of optical scans per second that the scanner performs internally. This should not be confused with the number of products that can be scanned per minute or per hour, which is the real issue that must be considered when evaluating a scanners performance. For example, Intermec says that a linear imager will have scanned the barcode three to five times by the time a laser has scanned it once. This shouldn't be taken to mean that a linear imager can scan three to five barcoded items in the time a laser takes to scan one barcoded item. Anyone who has ever compared a laser scanner to a linear imager knows that this isn't true. Real world comparisons by major retailers have consistently shown that a good laser scanner will provide unexcelled productivity.
It is also important to realize that the number of scans per second doesn't even give a good indication of how long it takes a scanner to read an individual barcode, let alone multiple barcodes. For example, we have measured how long the Intermec E1022 imaging scan engine (which was mentioned by Intermec several times in the Sept.12 article) takes to decode a single barcode. In our test, a stationary barcode was positioned in front of the engine before the scanning function of the engine was enabled. The timing measurement was started when the engine was enabled, (the equivalent of pulling the trigger on a hand held scanner) and ended when the engine indicated that the barcode had been decoded. The measured time varies a little depending on the distance between the scanner and the barcode, but the average time is within a millisecond or two of the time measured for Symbols recently introduced LS-2208 hand held laser scanner. When the barcode is moved during the measurement time the imaging engine slows down or stops decoding, but the laser doesn't.
Like most scanners, the E1022 also slows down somewhat as the distance between the barcode and the scanner is increased, until it ultimately stops decoding completely. Since the E1022 has less overall range than the LS-2208, it becomes considerably slower sooner as range is increased, and stops scanning completely at ranges where the LS-2208 still decodes quickly. The LS-2208 can therefore provide better overall productivity in scanning applications where it might be necessary to ocasionally scan at working ranges beyond that which the E1022 can provide.
Intermec goes on to say that they conduct scan-offs between lasers and linear imagers to see which one can scan a sheet of barcodes the fastest. It is my understanding that this test is conducted by asking the users to scan columns of barcodes affixed to a flat surface, starting at the top of the column and working down the column. This test doesn't test the motion sensitivity of a linear imager, as the scanner is always moving vertically from one barcode to the next, never horizontally, where imagers are far weaker. Therefore it should not be assumed that a scanner that performs well in this test will also work well in a real scanning environment. In addition, every barcode scanned during this test is at the same distance from the scanner, so the test doesn't differentiate between a scanner with more range and one with less range which can also be improtant in some situations. This test also fails to stress a scanners ability to read barcodes from various angles, which, as discussed in the next section, is another attribute of a scanner that needs to be optimized if a scanner is to provide the best possible productivity.
Angular Orientation Tolerance
In the earlier article, Symbol claimed that linear imagers have poorer tolerance for reading barcodes that are tilted with respect to the scanner. We were making the point that angular tolerance is important because the barcode will frequently not be sitting flat on a table in front of the operator. Barcodes can be on any side of a package and they can be rotated into any orientation on any side. A scanner that allows scanning from larger angles minimizes time spent aligning the scanner to the barcode, or moving the barcoded package to make the barcode easy to scan. Excellent angular tolerance also minimizes hand manipulation, wrist rotation etc. For these reasons, angular tolerance contributes to a scanners ability to provide superior productivity.
When we made the statement about the relative angular tolerance of lasers and imagers, we were speaking in the context of a comparison between the LS-4000 and the IT 5600. The LS-4000 and the newer LS-2208 both have significantly better angular tolerance than the IT 5600.
Intermec claims that their E1022 has angular tolerance that is as good as Symbol lasers they have tested. We have not measured the angular tolerance of the E1022, but we have measured the Intermec Vista, which we believe uses a similar engine. Our measurements show that, while the Vista does have better angular tolerance than the IT 5600, it is still not as good as the LS-4000, which was the subject of the earlier discussion.
Low Contrast
Symbol stated that laser scanners provide better performance on low contrast barcodes than linear imagers. Intermec responds that their linear imagers can read lower contrast barcodes than the vast majority of Symbol laser scanners. Clearly we need to define what we mean when discussing performance on low contrast barcodes.
Ideally, a scanner should be designed to be able to read low contrast barcodes so well that the user doesn't notice any difference in the behavior of the scanner when reading low contrast barcodes as compared to reading good ones. If the scanner acts the same on all barcodes the user can work the fastest. For example, if the scanners working range is severely reduced when reading low contrast barcodes, the user will find that they sometimes don't scan right away even though he is using the same scanning technique and scanning at the same distance as he has been using on all the other barcodes he has scanned. He will than either give up and manually enter the barcode data, or he will spend time hunting for the working range at which the low contrast barcode will read. Either way time is wasted and productivity is lost.
Our measurements have shown that the working range of the Intermec E1022 imaging engine on a low contrast UPC barcode (25%) is only around half of the range of an LS-4008i, which is the latest version of the LS-4000. In fact, the LS-4008i has more working range on a 25% contrast UPC (100% magnification) than the E1022 has on a mint quality high contrast UPC barcode. The result is that a user of the LS-4008i can seamlessly scan low contrast barcodes intermixed with good barcodes without breaking his stride.
When we originally made our statement about performance on low contrast barcodes we were responding to the comparison between the LS-4000 and the IT 5600. We have measured the working range of the IT 5600 on a 25% contrast barcode, and find that it is only about one third as large as the LS-4008i.
Intermec claims that their engine can read lower contrast barcodes than the majority of the laser scanners built by Symbol. Without running an exhaustive test we can't comment about that, but we can say that barcodes with contrasts lower than 25% are extremely rare outside of testing laboratories. In the real world a scanner that has good range on the more common levels of reduced contrast (down to 25%) will be easier and faster to use on a larger percentage of barcodes than one that loses a lot of range, even if that scanner might be able to read an even lower contrast barcode over an even more severely reduced range. In other words, a scanner that has superior performance on almost every barcode that will normally be encountered will provide better productivity in a real scanning environment than one that may be able to read barcodes that are very unlikely to be encountered, but doesn't read commonly encountered barcodes as well.
Size
Symbol commented that linear imagers are larger and heavier than laser scanners. Intermec disagrees with this and points to their E1022, which is around the same size as the Symbol SE-1200 laser engine, and they also point to their smaller EV 10 engine which is around the size of the Symbol SE-900 laser engine. As you recall, we made our comment about the relative size of lasers and imagers in the context of complete scanners, not unhoused engines.
If we want to talk about engines, however, we would like to point out that the laser engines with the equivalent sizes of the two Intermec engines are capable of providing significantly more working range than their imaging counterparts. In fact, our new SE-824 engine, which is around the same size as the Intermec EV 10, actually provides more working range than Intermec's larger E1022. In addition, the SE-824 uses a new signal processor that provides better performance on various badly printed barcodes than has been available before in a small laser engine. The point of this is that, for a particular size, a laser engine can still provide better performance than a linear imager. Alternatively, if a given level of performance is needed, a laser can supply that performance in a smaller package.
A smaller scanner is less tiring to use for exrtended periods of time, so a lasers ability to provide higher performance for a given size contributes to its ability to provide better productivity.
Laser Safety
In the Sept. 12 article, Intermec attempts to change their statement about laser safety, claiming that they said "consumers had some concerns about lasers". What they actually said in the earlier article was that in "cases involving scanning by consumers, laser technology may pose a safety problem". These are very different statements. Consumers might well continue having concerns about lasers if Intermec continues disseminating misinformation about laser safety. This doesn't mean, however, that there actually is a real safety hazard.
The fact is that lasers with the power levels used in hand held scanners are completely harmless. This statement is supported by, among other things, a test run by the Mayo Clinic in the year 2000. This test was designed specifically to investigate the question of the safety of lasers in the hands of consumers.
This test was prompted by the availability of laser pointers, which use lasers similar to the ones in hand held scanners, but which actually emit laser power that is significantly higher than any hand held laser scanner. In the test, the eyes of volunteers were exposed to energy levels from laser pointers that are literally thousands of times higher than what can happen with a hand held scanner. The test showed that no physiological damage to the eyes and no loss of any visual function occurred. This authoritative test surely must be given more credence than unsupported claims by Intermec's marketing director.
Laser Vs. Imager
Intermec states that lasers are best suited for applications requiring extended working ranges (up to 30 feet). While this is true, the implication in Intermec's statement is that linear imagers are better than lasers for scanning at closer ranges. In fact, a white paper published by Intermec explicitly makes the claim that CCD scanners are better when working at ranges of less than 18 inches.
Intermec's reasoning behind this claim fails to take into account any of the points we have made above. We have explained why a lasers insensitivity to motion, excellent angular tolerance, smaller size and superior working range on low contrast barcodes all add up to provide the best overall productivity. In general, the applications for hand held scanners where they are used at ranges of 18 inches or less are also the ones where the highest productivity is most important. Lasers are therefore the best choice even when scanning at ranges that can be achieved by linear imagers.
Intermec also states that they expect that the majority of customers will choose imaging technology in cases where either technology will work. While there are cases where either technology will work, lasers will still work better in many of them. Sometimes, however, productivity will have to take a lower priority than the cost of the scanner. Linear imagers have been less expensive than lasers, so they have often been purchased by users who didn't want to spend enough for a laser, even if they might have preferred to have one. The introduction of the new LS-2208, however, eliminates the need to sacrifice the important performance attributes of laser scanners for budgetary reasons. This scanner uses new technology that allows it to retain excellent laser performance while selling at a price that is actually lower than some linear imagers.
Intermec is correct when they state that the customers, and not the vendors, will ultimately decide which technology they prefer. Several major retailers have recently chosen to purchase laser scanners even though linear imagers were available to them, and were tested side by side with lasers in actual scanning environments. These sophisticated customers know that the best way to find out which scanner will work best for them, is to try it in their real application. They know that lab tests and data sheets are at best a poor indication of how a scanner will perform in the real world, so they make their buying decisions based on field tests, where laser scanners excel.
Bio - Ed Barkan
Edward Barkan presently holds the title of Senior Fellow at Symbol Technologies, where he has worked since 1977. Over that time period he has been the inventor and principal designer of many of the worlds most sucessful barcode scanning products. As examples of his significant contributions to the data capture industry, Mr. Barkan is the co-inventor of the hand held laser scanner and of the hand held scanning computer. He is also the creator of such ground breaking products as the LS-7000, which was the first comercially sucessful hand held laser scanner, the much immitated LS-9100, which set the standard for counter-top projection scanners and the SE-1200 and SE-900 laser scanning engines which are used in millions of portable scanning computers and hand held scanners world wide.
Mr. Barkan holds 150 patents in the fields of barcode scanning and related areas. |
