Sean Fenske, Editor-in-Chief04.25.24
Welcome to another #ThrowbackThursday column in which I revisit Editor’s Letters from my 25 years covering this fantastic industry. I hope you’re enjoying reading them (if this is your first, check out the others by clicking here) as much as I am writing them.
For this next entry, I’m combining two Letters into one as one doesn’t have enough “meat” on it to be a stand-alone presentation. That Letter from March 2000 issue of Medical Design Technology (MDT) spoke of the importance of “Building Your Internet Library.”
In the early-ish days of the World Wide Web, internet searching/browsing was not nearly as commonplace as it is today. Sure, many people were connected by 2000, but with dial-up still being the primary method for accessing the web, it was not nearly as simple as it is today. You had to wait for pages to load and had to hope no one picked up the phone line as you were attempting to gather information.
Still, I found it interesting that I noted in a search for “medical device manufacturing,” results tallied 1,633,131 (that same search today generated 373 million results). That said, in this letter, I was attempting to highlight several “must-have” bookmarks for medical device manufacturing professionals. These sites were for FDA, ISO, AAMI, HIMA (previous name of AdvaMed), MDSS.com and MedicalDevices.gov.uk (for European regulatory info), and MDT’s home page (which is no longer a thing). To go into any further depth on this topic would seem silly so we’ll move onto the next Letter.
A recent Letter discussed machine vision; this next one brought up machine hearing. The April 2000 MDT Editor Letter was titled “Are Your Machines Listening to You?” (with the italic in the headline added for extra emphasis to induce amazement; does it still work today?).
Imagine being able to control a machine without the use of a keypad, remote control, or touch screen, but instead, simply stating orders into a microphone. This option may not be as far off in the future as you may think. Voice recognition technology is currently available and it is only a matter of time before we see it in both medical device manufacturing processes and their products.
Funny enough, medical device manufacturing is one of the few places I don’t see voice-based technology being used all that much (I certainly encourage anyone using it on the production floor to share their story with me). But it’s everywhere else. We have voice-based devices in our homes in the form of Amazon’s Alexa and Google’s Home or Nest, in our pockets in the form of our smartphones, in our cars with voice-activated options, and even on our TV remotes. Surgeons are able to leverage it with certain devices, such as displays to avoid having to touch them. There are countless other applications where this technology has been introduced, but as far as I know, the production floor is not a common one.
Voice recognition technology is now being used by doctors in the medical field primarily for dictation, which eliminates transcribing lengthy notes from various medical procedures. However, the concept is able to be adapted to many other applications, benefiting the medical industry.
Having used transcription technology in the 2000s, I hope a medical professional was reviewing those results carefully as it was not nearly as reliable as it is today.
One company (VRS LLC Medical Charting Systems) has created a software package, eliminating the traditional paper chart while moving toward a comprehensive electronic listing that makes available all patient information quickly and efficiently. Although the software system has a combination of touch screen and voice recognition technology, it will not be long before the need for manual input is completely eliminated and doctors will be able to record patient data through speech only.
First, as far as I can tell, VRS didn’t make it. They may have been acquired but a listing I found for them from 2002 provides a now-defunct URL. Next, I can’t speak to your experience, but my doctor visits today involve a physician and a laptop with no voice recognition taking place. I go back to accuracy and wonder if that’s still a concern today.
While data acquisition is certainly a very viable aspect of future technology, another company (Sensory Inc.) has a variety of advanced systems, which could easily be incorporated into medical device technology and the machines utilized to manufacture it.
This company IS still around, has won several awards, and seems to be working with leading technology innovators. The company maintains an emphasis on voice-based technologies but has since added biometrics and artificial intelligence to its portfolio. Medical is listed under industries but it’s grouped in with banking and enterprise, so I’m not sure how much of a market medtech is for the organization.
According to my Letter from 2000, capabilities then included:
Speaker-Dependent Speech Recognition enables the operator to utilize vocabularies of over 100 words. This type of system would be desirable in an assembly situation where the same person controls a specific workstation, at which a variety of commands are required. However, if there was a need to have multiple persons at this station, the voice commands can be saved to a floppy disk or other form of data storage device to allow them to be interchanged quickly and easily.
I would have to look hard for another paragraph that illustrates just how far a technology has evolved over a period of time. I feel like this is a story my grandfather is telling me about the olden days of computers.
Speaker-Independent Speech Recognition does not require any training for the end-user and is programmed for up to 14 commands. This system is advantageous for any device or machine that might be utilized by a variety of personnel but does not require an extensive array of commands. Utilization in certain types of testing equipment and instrumentation would be some possible applications.
All of 14 commands, huh? Boy, this tech was really built to impress. The fact there are different levels of voice recognition capabilities alone is surprising to hear.
Speaker Verification requires the user to speak into the system in order to gain security access. This type of technology would be potentially useful in IV equipment set-up to automatically dispense medicine to patients, whereas doctors would have to gain security clearance through voice recognition in order to alter a dosage.
This type of capability is the first that’s reasonably useful for today. Although, even my home voice-based device recognizes when I give it a command versus another family member. So again, these solutions and more are all available today and in technology that’s much more common that when I wrote this piece.
There is a variety of other voice interface systems currently available and future technologies promise to bring us more. It is up to you, the engineers and technical personnel, to develop effective and useful methods with which to integrate this technology so as to enhance the efficiency of a specific device.
Fortunately, they’ve already been doing this for years and that integration has been successful.
This is the type of innovation that design engineers and technical professionals need to explore and research while developing medical devices, bringing us one step closer to what is perceived as “future technology.”
ESL is still around, but just how successful or popular its voice recognition features were for testing systems is unknown (should the company be reading this and would like to weigh in, I’m happy to share your comments). The premise is sound though; a test engineer needs use of his/her hands to prepare samples or note data while still maintaining the ability to conduct a test with voice commands. However, now those commands are likely pre-programmed and once a single button is pushed, the test system runs through the required actions until complete when the next sample is loaded. Step-by-step commands throughout the test are likely not necessary.
For this next entry, I’m combining two Letters into one as one doesn’t have enough “meat” on it to be a stand-alone presentation. That Letter from March 2000 issue of Medical Design Technology (MDT) spoke of the importance of “Building Your Internet Library.”
In the early-ish days of the World Wide Web, internet searching/browsing was not nearly as commonplace as it is today. Sure, many people were connected by 2000, but with dial-up still being the primary method for accessing the web, it was not nearly as simple as it is today. You had to wait for pages to load and had to hope no one picked up the phone line as you were attempting to gather information.
Still, I found it interesting that I noted in a search for “medical device manufacturing,” results tallied 1,633,131 (that same search today generated 373 million results). That said, in this letter, I was attempting to highlight several “must-have” bookmarks for medical device manufacturing professionals. These sites were for FDA, ISO, AAMI, HIMA (previous name of AdvaMed), MDSS.com and MedicalDevices.gov.uk (for European regulatory info), and MDT’s home page (which is no longer a thing). To go into any further depth on this topic would seem silly so we’ll move onto the next Letter.
A recent Letter discussed machine vision; this next one brought up machine hearing. The April 2000 MDT Editor Letter was titled “Are Your Machines Listening to You?” (with the italic in the headline added for extra emphasis to induce amazement; does it still work today?).
Imagine being able to control a machine without the use of a keypad, remote control, or touch screen, but instead, simply stating orders into a microphone. This option may not be as far off in the future as you may think. Voice recognition technology is currently available and it is only a matter of time before we see it in both medical device manufacturing processes and their products.
Funny enough, medical device manufacturing is one of the few places I don’t see voice-based technology being used all that much (I certainly encourage anyone using it on the production floor to share their story with me). But it’s everywhere else. We have voice-based devices in our homes in the form of Amazon’s Alexa and Google’s Home or Nest, in our pockets in the form of our smartphones, in our cars with voice-activated options, and even on our TV remotes. Surgeons are able to leverage it with certain devices, such as displays to avoid having to touch them. There are countless other applications where this technology has been introduced, but as far as I know, the production floor is not a common one.
Voice recognition technology is now being used by doctors in the medical field primarily for dictation, which eliminates transcribing lengthy notes from various medical procedures. However, the concept is able to be adapted to many other applications, benefiting the medical industry.
Having used transcription technology in the 2000s, I hope a medical professional was reviewing those results carefully as it was not nearly as reliable as it is today.
One company (VRS LLC Medical Charting Systems) has created a software package, eliminating the traditional paper chart while moving toward a comprehensive electronic listing that makes available all patient information quickly and efficiently. Although the software system has a combination of touch screen and voice recognition technology, it will not be long before the need for manual input is completely eliminated and doctors will be able to record patient data through speech only.
First, as far as I can tell, VRS didn’t make it. They may have been acquired but a listing I found for them from 2002 provides a now-defunct URL. Next, I can’t speak to your experience, but my doctor visits today involve a physician and a laptop with no voice recognition taking place. I go back to accuracy and wonder if that’s still a concern today.
While data acquisition is certainly a very viable aspect of future technology, another company (Sensory Inc.) has a variety of advanced systems, which could easily be incorporated into medical device technology and the machines utilized to manufacture it.
This company IS still around, has won several awards, and seems to be working with leading technology innovators. The company maintains an emphasis on voice-based technologies but has since added biometrics and artificial intelligence to its portfolio. Medical is listed under industries but it’s grouped in with banking and enterprise, so I’m not sure how much of a market medtech is for the organization.
According to my Letter from 2000, capabilities then included:
Speaker-Dependent Speech Recognition enables the operator to utilize vocabularies of over 100 words. This type of system would be desirable in an assembly situation where the same person controls a specific workstation, at which a variety of commands are required. However, if there was a need to have multiple persons at this station, the voice commands can be saved to a floppy disk or other form of data storage device to allow them to be interchanged quickly and easily.
I would have to look hard for another paragraph that illustrates just how far a technology has evolved over a period of time. I feel like this is a story my grandfather is telling me about the olden days of computers.
Speaker-Independent Speech Recognition does not require any training for the end-user and is programmed for up to 14 commands. This system is advantageous for any device or machine that might be utilized by a variety of personnel but does not require an extensive array of commands. Utilization in certain types of testing equipment and instrumentation would be some possible applications.
All of 14 commands, huh? Boy, this tech was really built to impress. The fact there are different levels of voice recognition capabilities alone is surprising to hear.
Speaker Verification requires the user to speak into the system in order to gain security access. This type of technology would be potentially useful in IV equipment set-up to automatically dispense medicine to patients, whereas doctors would have to gain security clearance through voice recognition in order to alter a dosage.
This type of capability is the first that’s reasonably useful for today. Although, even my home voice-based device recognizes when I give it a command versus another family member. So again, these solutions and more are all available today and in technology that’s much more common that when I wrote this piece.
There is a variety of other voice interface systems currently available and future technologies promise to bring us more. It is up to you, the engineers and technical personnel, to develop effective and useful methods with which to integrate this technology so as to enhance the efficiency of a specific device.
Fortunately, they’ve already been doing this for years and that integration has been successful.
A Success Story
As I stated earlier, this voice recognition can easily be incorporated into medical device manufacturing equipment and their products to make them more user-friendly. One company (Electro Standards Laboratories) has done exactly that with its ESL Test II Digital Test system. This machine performs industry standard material testing, specifically tensile, compression, and peel. It is among the first pieces of equipment to utilize voice recognition technology for hands-free operation, enabling the user to collect data or prepare additional test samples, while still maintaining complete control of the machine.This is the type of innovation that design engineers and technical professionals need to explore and research while developing medical devices, bringing us one step closer to what is perceived as “future technology.”
ESL is still around, but just how successful or popular its voice recognition features were for testing systems is unknown (should the company be reading this and would like to weigh in, I’m happy to share your comments). The premise is sound though; a test engineer needs use of his/her hands to prepare samples or note data while still maintaining the ability to conduct a test with voice commands. However, now those commands are likely pre-programmed and once a single button is pushed, the test system runs through the required actions until complete when the next sample is loaded. Step-by-step commands throughout the test are likely not necessary.