>> Good morning, everybody, good afternoon who is in Europe and good evening for attendees from Australia. My name is Emanuele Cauda and I'm very excited to have all of you at this webinar today titled Transformation of Health and Safety Sensors Data into Information and Action: Real-World Examples. I'm very excited about this webinar today. I've worked with my colleague, Emily Ash and Valerie Cog now for a few months on the preparation of this webinar and I'm very excited that we got to the day of going live. I also want to immediately thank Kiana Harper here at NIOSH, who has all the tools to help us out with the webinar from the Zoom perspective, and hopefully it's going to be a smooth webinar from that perspective. This webinar is organized by the Center for Direct Reading and Sensor Technologies.
As a center, you can read more following the QR code. We have a mission to coordinate research and to develop recommendations on the use of the 21st century technologies in occupational health and safety. I really need to thank Valerie because now we also have a mailbox sensor@cdc.gov, if anybody wants to interact with the center. We have been involved in the idea of the transformation of sensor data into information and knowledge for a few years.
Exactly one year ago, we had an internal workshop as NIOSH with two international partners and collaborators, TNO in the Netherlands and HSC in the United Kingdom. And we have a blog on that to describe the finding of that internal workshop. But more in general, we've been involved in the idea of bringing the profession of industrial hygiene and occupational hygiene in the future to what we consider industrial hygiene 4.0. You can read that in a recent article in The Synergist, which is the journal of the American Industrial Hygiene Association, AIHA. So today's webinar, I'm very excited about the fact that three professionals have agreed to present their real-world studies. And every webinar on sensors generally is really heavy on technology.
We're going to have quite a bit of technology information, especially from the first presenter, but I'm really happy that all of them agreed to focus primarily on the theoretical idea of the transformation of data from sensors for health and safety or occupational hygiene into what can be applicable information and new knowledge. So we're going to start with some updates from Professor William Mills on their effort of creating a low-cost network, actually an environmental quality monitoring and control system. Then, we're going to hear from Norway, from Ase Austigard on the interaction of workers and gas sensor monitors. And to finish with Corinne Balcean from Teck in Canada, Teck Resources, on the operationalization of real-time particulate monitoring for occupational hygiene. It's a really nice progression.
Each presenter and panelist is going to have around 20 minutes to present their contribution, and then we're going to have a question-and-answer moment after each session, after each contribution. And then at the end of the three contributions, we're going to have a roundtable discussion. I really would like to encourage you to use the Q&A box in the Zoom webinar platform to submit your questions. We are going to collect them during each contribution and convey to the panelists, and also at the end for the roundtable discussion. With this, I'm done with my introductory remarks, and I would like to pass the microphone to my colleague Dr. Emily Ash for the introduction
of the first panelist. Please, Emily. >> Thank you, Emanuele.
So yes, it's my pleasure to introduce our first speaker for the day, Dr. William Mills, who is an associate professor in the Department of Engineering Technology at Northern Illinois University, where he teaches and has 40 years of experience in environmental health and safety. His laboratory research has included extensive work with sensor systems for indoor environmental quality monitoring and management. Dr. Mills is an active member
of the Real-Time Detection Systems Committee in AIHA. He holds a PhD in environmental and occupational health sciences from the School of Public Health at the University of Illinois at Chicago. So take it away, Dr. Mills.
>> Thank you very much there. Good morning, everybody, or afternoon and evening, as they say. Thank you for joining in today, and I'm just going to get my slides started.
While I'm doing that, the slide deck that I have provided has far more slides than I'm going to cover in the slide. That was on purpose. So I'm going to skip over, as Dr. Cauda or Emanuele mentioned, we agreed to try to make it a little bit higher level on what do you do with the data.
I need to acknowledge my colleague Dr. Kevin Martin, who's the co-founder with me of the Building Energy Efficiency, Ergonomics, and Management Lab, where most of this work is done out of, or BEEEM lab at NIU, and Matthew Bikun and Justin Cathey have both been graduate students, and some of the work you're seeing is, in fact, their work. As a government state university, we have to put the standard disclaimer. You can all read it.
So we're starting, and we've been working since 2014 there. This is what we called our BEEEM Building Control System in 2016, it's a prototype system, really just did control. It didn't do any logging, it was like a dumb control system, but it has been published, focused around Raspberry Pi technology and a variety of different sensors. So I think everybody has realized, the last two years, COVID, so in the fall of 2021 when we were returning to in-class instruction at Northern Illinois University, there was a lot of concern from faculty, especially -- Northern Illinois has been around since 1895, was when it was first established, and many of the buildings are in the early part of the century. And when we were returning, there was no upgrade on the HVAC going in.
What had gone on is that they were HEPA units bought by the university for classrooms that were going to have instruction, however, they were industrial units, they were very loud, had two speed settings, but loud and louder. So there were a number of issues concerned by faculty and students, staff. One was, some of the questions we had, a number of the units that should be operating in any of the spaces, where should it be placed? And is a room, even with the HEPA properly ventilated under reduced occupancy or full occupancy? And so this is sort of the background on how we got going on the work I'm going to describe to you. First thing I want to do is a shout out.
If people have not heard about it yet, Emanuele and myself, Steve, John, and several others have been working diligently for almost a year now on the AIHA alarm settings white paper. And out of that, we developed a five-step process, which is shown here. You can all read, and you'll also see a diagram of it coming up.
But I want you to be aware of this process, because we're going to go back, because we did validate it by using the work I'm going to describe. So one of the first things that comes out of that is determining what the purpose of your monitoring is. And we came up with, in conjunction with, I think it was either HSE, I believe this is mostly from HSE -- from gas monitoring looked at, they really only had two: safety monitoring and health monitoring. And we added the C and D, which is combination or other purposes, which is really more what this type of project falls under.
In our initial evaluation, we looked at indoor air quality parameters, and what we wanted to look at to describe whether we felt we had adequate ventilation for protection for everybody in the classroom. Typical ones, you look at temperature, relative humidity, CO2, PM2.5, and volatile organic compounds. And there are a couple additional parameters that we felt could be helpful, including light measurements, noise measurements, and particle size distribution. And we evolved with those additional parameters more to the proper term being indoor environmental quality rather than indoor air quality.
First thing we looked at, at the time, and why we ended up doing the work we did is we did look around at what some of the low-cost air quality monitors are, and these are just some representation. But there were a number of drawbacks, maybe not drawbacks, a number of considerations that we felt made it important to look at some alternatives. One of the things is the need for all of these to have internet connectivity in order to really be looking at the data. Other than one of them here, there's no accessible local data storage and no real-time plotting of that. In all these cases, they are trying to keep the data on the cloud. There's issues with ownership of the data.
Different download intervals can sometimes be a problem. What they call API, Application Programming Interface, which allows you to go out to the cloud and bring the data in usually. Many of them charge ongoing for data management, and most of them did not have all the parameters that we really wanted to cover.
So we've been working, as I said, for many years at NIU, and we've worked prior years. You can see in the upper right-hand corner, I think my mouse does show, I hope so. And in the upper right, with the single dollar sign, are some low-cost sensors that we've worked with. These are around $60 plus or minus. We've compared them to a number of different of our typical industrial hygiene commercial instruments in the bottom left there. I'm not going to go over all of them, but just to show you how well those low-cost aerosol sensors can do.
Here's a 0.3 micron channel of three different sensors, plant tower sensors versus a particles plus optical particle counter in the 0.3 micron range. And the conclusion from looking at this would be that those low-cost sensors are fit for purpose to make decisions with. I talked about fall 2021.
Let's put this in perspective. This is with what's called a Q-Track XP, and I'm able to do six particle channels on there and also convert that to PM2.5 readings, which most people are familiar with. And what I want to really show you in both cases, what you see is a gradual rise in a room that was at about half occupancy, and everybody was wearing masks, but the HEPA filters were not running because it was just too loud. But you can see these sensors can see the buildup of the respiratory aerosols. I'm not saying this is major amounts.
If you look at the PM2.5 or the particle counts, those are not large readings, but they can catch respiratory aerosols. And for the low-cost carbon dioxide, the one I'm going to talk about the most is a Sensirion, what's called an SCD30, and that's the red dots on here. And we ran that with a number of those low-cost sensors, including some of the commercial ones, and an EVM 7, which is about a $10,000 instrument. So that's a $60 CO2 sensor versus a $10,000 instrument that includes CO2.
So I think, again, the conclusion, these sensors are fit for the intended purpose. So we came up with a prototype, and you'll notice I say 1.0, 1.1, and 1.2, and the issue here, those dotted lines are ones that were not there in 1.0.
And the dotted line in the bottom right sensor, you'll see that we also have a Sensirion SPS 30 that we've put in some of our IEQMs. We added a smart outlet to monitor when the HEPA filter was actually on, and there is now what's called a photoacoustic spectroscopy, and that's that Sensirion SCD 40 that we've also been working on as another alternative. But the general IEQM is shown here with the light sensor, the CO2 temperature and relative humidity, and the SCD30, a microphone, and a particle counter, all of this going back to a Raspberry Pi for logging the information and displaying the information. One thing that turned out to give you -- in order to make decisions with the data, the actual question of averaging time, and it turns out that we do need to use five-minute averaging time, that gives us the best optimum there.
The other thing we looked at for why we wanted to use particle counts, and we found the particle count 0.3 is, you can see, it's pretty constant in terms of the relative fraction across all the work. So this is some real-world data in two different classrooms.
The one on the left is a relatively large classroom with up to 40 students, and the scales are different. The one on the right is actually a reasonably large classroom, and so on the right. But what we found is that's got only radiators and air conditioners, and it gets much higher CO2 levels than the one on the left. The other thing from our monitoring that I can talk about with people who have questions, outdoor air matters for indoor air.
And this is actually showing our indoor air on the IEQM compared to the local purple air using similar sensors in the DeKalb area, and you can see that there is an impact. Another factor that's turning out to be really important in using sensors is your data management. It's very easy now to generate a lot of data for a lot of different parameters, but what do you do with it? And so there are a number of things that you can do to help with this. First of all, try to come up with some standardization on how you're going to work with all your different sensors.
We use Python and C++ programming language. Your data management system, especially with this type of data, having a database that can handle time series data. We use Influx. There's also SQLite, but whatever you use, having some sort of data management system that you can then work further with. And then you have to have a way of doing data analysis, and we did a lot of work. We implemented what's called R Project, which is open source, and then using all this data to come up with some actual decisions.
And that's really where the initial sensor, looking at indoor environmental quality, we made the jump after I showed you some of that data to actually building management. So we took the base monitoring, and we've added several types of control into this. Now this is a prototype system. It is actually working in a room, but it is more like -- it's actually a computer room that we've got this set up and controlling. So that's why you'll see in the upper right, there's what we call the AC Pi, the air conditioning Pi, where it's actually able to monitor and control the air conditioning, heating, and air circulation unit. And we hacked into the infrared remote with a Pi hat that does infrared control now.
So we can change the cooling and the heating and the fan speed with that Pi, and the power we can either turn it on or we can also monitor the power with that smart plug. And one of the things with that design, we were having some problems with ice buildup. So we actually implemented in the middle there a low, it's called a NoIR, noir, Pi camera, Low-cost camera with its own Pi. And we did open CV, that's computer vision, and we wrote an algorithm where we could cycle the computer so that -- because we could see the compressor ice buildup.
So there's another decision making type of thing. And on the very left, we put in a HEPA, that's a Honeywell AirGenius 6. It normally uses Bluetooth, and so we hacked the Bluetooth channels and made our own set of Bluetooth controls, and we can go through five, I think it's got five different speeds, and we can rotate it, et cetera.
And we can either turn it on and/or monitor the power to make sure it is in fact running through the smart outlet. And in the bottom left are all the sensors you've already seen. One other thing that we added there, if you look, the ESP microcontroller, which is its own little board and allows us to move these -- we have seven, or I think we had six to seven of those DHT22 temperature sensors around the room, and there's a number of different communication protocols I'm not going to go into. What I really do want to talk about, though, is how using this system, we were able to validate that five-step process. Let's walk through this with what you've actually just seen. The monitoring of the room environment, initially we went into it to just look at monitoring what was going on with CO2, the aerosol temperature, relativity, and noise and light.
The objective, we chose D, which we were doing research, but we were also looking at possible exposure assessment, and we've added control here. If you then look through your selection of the sensors and your fitness for purpose, you'll see we chose a CO2 with non-dispersive infrared, and in the 1.2 version of the IEQM, we actually now have a photo-acoustic spectrometer in there.
And the SCD30 included temperature and relative humidity. For aerosols, that was the initial selection of one of the plant tower OPC units, and in 1.1, 1.2, we looked at -- we've also included the Sensirion SPS30. For noise, we selected Dayton iMM-6, which is -- Dahosh has done a lot of work for that for 29 bucks, and you can hook that on its own also with a smartphone.
We've tested against $3,500 octave band analyzers, and it's performed quite well. The light sensor, the basic one we use only gives you lux, but we have also done work with a Nanolambda XL500 spectrometer, and also an MP32M Nanolambda spectrometer, and that allows us the spectral power distribution, so we can take into account spectral shifting and LED lights. For power, we added that, really that's version 1.2, with a smart outlet, which either allows you to turn things on and off, or monitor, for example, the HEPA, what speed it's on to go along with -- so when you're looking at the air quality, one of the options you can look at is what's the impact of the HEPA speed. And the temperature and relative humidity, that was the DHT-22 you saw with the ESP boards. Where we think -- so it's this step three and step four that we think the white paper is going to have its major contributions, and so I'm showing here the set points that we came up with for the IEQM and management system, and so first of all, CO2, we basically came up with three regimes, less than 750 we considered green, 750 to 1,000 is hey, something's going on.
And certainly you don't want to get over 2,000 part per million. If you recall back to my slides on those two rooms, and we did a number of rooms, but I'm just using those two to illustrate, that room that was on the right that only had radiator and ventilation, no active air, with only 12 people in, even though it was allowed to take up to 50, we almost got up to 2,000 part per million sometimes in classes, and in most of the classes we were over 1,000 part per million. So that was one of the first indicators, you're just not getting enough fresh air in there to ventilate. For PM2.5 or the other channels, for PM2.5 on a health base, we found that a background of 5 to 10 was pretty typical for most of these classrooms when it was unoccupied. We did not really see in any of the classrooms I was monitoring major increases in the PM2.5, but there are health base standards
provided there. The PC.3 we think is one of the biggest advantages of going with optical particle counters. It's a very sensitive channel.
You could see when I showed that room, it was less jerky and I think the PC -- having the particle counts in different bins, the PC.3 was by far, the 0.3 micron has the most. But it's a new parameter and it's going to be dependent on every space. So you need to look at your backgrounds and your indoor background, but also your effective outdoors. So that's why I have a question mark, because we're still trying to work through what we think would be a control on there.
We do have programmed in for the HEPA and for the ventilation, you saw that we had the ability to recirculate the air net for CO2 and PM2.5. And then for temperature control, we basically want to be above 20 degrees Celsius in winter and try to be below 25 to 27 degrees Celsius in summer. And then stakeholder communications is the final part on this.
So University of Professional Illinois, which is a faculty unit, I've talked to them about what we found from this work. Another example of this is this talk itself. And we have a much more detailed case study for this that will be coming out in the AIHA white paper.
Now the other thing I want to emphasize on this one is notice the do loop. So if you think of a plan, do, check, act, it's actually built into here. That's the dotted line on the very bottom there. And you need to revisit on a regular basis when change occurs. So I'm going to go very quickly here. A lot of improvement, better variety of sensors.
I think we had much better data management, control logic, and broader applicability. Number of lessons learned. First thing I would emphasize to people is every space is different. We have 70 different monitor systems set up. Every space is different. It's much better to collect data in a space or a couple places in a space.
And the monitoring during occupancy is important. Or if you really wanted to get some idea beforehand, you could use dry ice to challenge and see how the CO2 is dissipated, and you can get actionable data for this. We have a number of future improvements, and I'm going to speed through here to get everybody else out here. So I do want to acknowledge funding that we've received over the years from the NIU College of Engineering and Engineering Technology.
We've received funding mostly in the way of buy-in of sensors and various instruments and helping with calibration from NIOSH, which we're grateful for. The University of Professional Illinois also provided funding both for my work and also brought a number of others, when I talked about those 70-sensor systems. Illinois Federation of Teachers and American Federation of Teachers have supported my work during COVID, especially for schools around the country. The American Society of Safety Professionals provides student scholarship funding for both Matthew Bikun and Justin Cathey, and some of the work you saw has been from other students.
Every year they have to do what we call senior design, and some of the work you saw is from that work. And with that, I'm going to turn it over to others. I'm happy to discuss it further, and in your slide deck, I think I covered about 25. I think I got about 60 slides in there, which has much more background for people. And feel free to reach out and contact me, I've got my email address here. All right, I turn it over to the next speaker.
>> Thank you so much, Dr. Mills. So we'll just -- there are some questions for you in the chat, and we'll let you look at those, but we'll ask one of them before we move on to our next speaker. So we have one comment and question that says, great work, particularly the five-step process and practical guidance. Sensor location and number of sensors have continued to be a challenge during field applications. Do you have any lessons learned to share other than what you already shared in the presentation, and will this be catered for in the five-step process? >> Yes and yes.
So can we bring that question back up at the end? >> Sure. >> I think that might be the best use of this, because it's bigger than just this one step. >> Okay. >> I do have some thoughts on that. >> That sounds great. You'll have some time to reflect during our other presentations, and yeah, there are some things in the chat that you can also -- not the chat, the Q&A that you can look at as well, and we can return to those questions too at the end if we have some time.
So thank you so much. And now we will move on to our next presenter, who is Ase Dalseth Austigard. Ase has been a certified occupational hygienist at the municipality of Trondheim since 2000, but she holds almost 30 years of experience as an occupational hygienist. She has a Master of Science from the Norwegian Institute of Technology, and is a PhD candidate at the Norwegian University of Science and Technology.
She was a faculty leader of the Norwegian Occupational Hygiene Association from 2010 to 2016, and a board member from 2002 to 2004, and served as the NYF Vice President from 2010 to 2016. So welcome, Ase. >> Thank you. Thank you very much for the invitation, it's a pleasure. Today I want to talk about our experiences on how equipment interacts with workers. These are some overall numbers for the water and wastewater department in Trondheim.
In a Norwegian scale, Trondheim is a large city, and our two wastewater plants are among the fifth and sixth largest in Norway, although they are small internationally, we see that. We have used personal sensors since approximately 2012, and have now also collected and worked with presentations of our data from approximately 2015. I'm going to talk about interactions between humans and instruments, and this starts when we realize we need an alarm sensor for safety, and what considerations do we do? And how will the workers interact with this equipment? I have learned from a colleague that this is a part of organizational psychology. I'm not in that part, but I present our observations and experiences from Trondheim Municipality. And we have observed a range of considerations and interactions that are present, probably more than these two, and these considerations and interactions fall into different groups, and I try to present them group by group.
And I'll start here. This is the special interest of the occupational hygienist, and that's normally where I come into the evaluation, how to choose. And the first one is, what shall we measure? What gases or particles do we want to have numbers on? And are there anything in the surroundings that we don't measure, that we need to be aware of? I need something on range, level of detection, resolution, and I've found out sometimes that I need to check the accuracy just as much, because I too often experience that the accuracy is poorer than the resolution, and I reckon others have the same experience.
And it's quite annoying when you realize that too late. If you want to use data later, not only use the instrument as an alarm tool, there is real interest, what is the time to reach 90% of the true concentration, T90 often noted in specifications. And also the storage capacity and logging interval, can you change it, or is it only a fixed time interval? And all these considerations are much the same as when you choose other measurement equipment as occupational hygienist.
These considerations, in addition to the alarm settings for light and vibration, the workers don't care much about this. It's something that's just there, but it doesn't bother them in any way. And adjustability of alarm settings might interest them, but mostly they don't care about that one either. Then there are some that workers, they don't care as long as it works. And that's the docking and charging functionalities. Docking and charging, especially charging, is necessary for the equipment to work.
Docking is necessary if you want to use the data later. And how that is done differs between the instruments and it differs with what you invest in infrastructure. I've seen many times that the calibration and bump test is not done. And especially if the infrastructure of dockings are not present. So this docking and charging, it takes time.
And if the equipment is not easily accessible, the workers don't do these parts. That's where we get our first mismatch if we want to use data. And in the beginning of our use, we didn't get much data, because the docking procedure said you should bump test the instrument and then you should press for data transfer and for calibration if that was necessary. But the workers, they were happy when the first green light came.
Then I took the instrument off and I didn't get any data. And the employees didn't get any data to evaluate how the exposure was. So when we realized this and we managed to change the docking procedure, we made an automated docking procedure. So they just slip it in and turn it on and then they can go to lunch or home.
And the next one, when the lights are green, can take it out, put it in the docking station, which must be nearby, come back to that, and slip in their own and they can go. But if the charging station is not close to the docking station, the first one's equipment is just left on the shelf and not charged to the next day. There are some instruments today that don't need any charging. I've not checked much of them, but I reckon there's not much other functionality. When you have a three-year battery life, battery use without charging, if you want to have it as a safety tool. So I still think the charging part is essential for getting good data over a long time.
And then we have a lot of things that the workers do care about and have meanings about, and that will affect their handling of the equipment. First one is weight and size. It must not be too heavy. If it's too heavy, they won't use it continuously, but leave it somewhere till they think that they need it.
Or if it's too large, they take it off because it's in the way when they are working. They also need a secure way of fastening the equipment to the clothes. I lost at least one, and that's not a good feeling. It must be easy to operate, preferably only a button, and that means on/off. Especially if you want to log any data, if the workers have to go to start procedure with more than one button and choose logging, you probably lose a lot of data. Alarm settings, I've talked about light and vibration before, but sound, that's one of the things that the workers care about.
If that sound is annoying or irritating or they can't close it off to work on, if that's part of the work, that might be one thing that makes them shut off the equipment. What does an alarm mean? Does it mean you have to evacuate? Then it's okay. Then you don't have the possibility to turn off an alarm. But what if it only means you have to be aware of, and you are dressed with a PPE, you are going to continue to do the work? If you don't have the possibility to turn off an alarm, you probably put the sensor somewhere that's not very good for the use as an alarm equipment. One challenge is peak levels.
Does an alarm go off at any peak level, no matter how short that peak is? How high must that peak then be? I'm working on H2S, hydrogen sulfide. In my work, also short peaks matter. But what about other gases? So we have to consider what are you measuring on and what time span should the peak tell about. But there is a job to do on distinguishing between different alarms. My instrument has four alarm settings that can be a TVA alarm, a cell alarm, a peak alarm.
So if there are different sounds, that might be okay, but if there are the same sound, no regard of the alarm type, that might be counterproductive. Workers also care much about GPS tracking. It's become more normal and raises some privacy issues. It opens, of course, for possibility of false sensors and help if you work alone.
And some of them also have microphones that can be connected to an alarm center, possibly also with a two-way alarm, two-way communication, and that's good. But it also opens for opportunities on surveillance, a fast drive, where you are, should be, and customer demands. I want to know where my worker is, when my workers show up, why isn't he here yet? And how should this be done without being intruding? And microphone, does it record the sound? I even get that question when I'm out doing noise measurements. So it is, in fact, a question that the workers care about.
If the equipment in any way are annoying or irritating to the workers, or they feel under surveillance, they will in some way act against the wishes of good data collection and functionality of the equipment. And the simple thing is where they do put the equipment when at lunch matters. Do they put it in their wardrobe, or do they turn it off, or do they put it in an exposed area? Those things are important to know. And if they turn it off, then targeted average alarms often start a new evaluation to the TVA when they turn it on again. So they might lose some of the alarms that they should have had. How they use it is also affected by organizational aspects and expectations from the boss.
It matters whether your boss expects you to turn it on when you come to work and use it all day, or if he expects you to use it if you are exposed. And our experience with H2S is that workers' expectation on which parts of the work are exposed is not good enough. They say this is not exposed, and we still measure H2S. So unexpected things happen, and then the equipment might be off if they only use it and they think they are exposed.
So we see that trust is an important issue if you want people to act in a specific way. If they don't trust you or don't like your decisions, they will resist in some way. And this is the psychological part of the equipment use. Regulations differ between nations and countries.
In Norway and in Scandinavia, we have what we call a three-part model, and it stands quite strong. It's a cooperation between the unions, the employers, and the lawmakers for the best of all parts of working life. And the law also demands that this is a cooperation that operates on the local scale. And if this cooperation works, it builds trust between the participants. You can use privacy as an example, as this stands strong also at work. If you expect a delivery from a worker and want to know when he or she comes, you might be given how far away this person is, but not the position.
This means that the employer uses positioning data, but they don't give out who you expect is at the moment. But you can be given a time for when you can expect him or her to be at your house or your place. In other instances, real time can be okay. And internally, we use it in the ski track preparation machines. So you can see the actual position now, but also how long is it since this or that ski track was prepped. In the sewage department, it is used, for instance, for positioning of TV cam monitoring of the pipes, but that's not used real time.
But it's there as technical data for inspection and so on. But what's important is that if a decision on tracking is made without involvement of the employee representative, at least in Norway, this can be deemed as unlawful. So that's an important part to follow up. So what to consider when you choose an alarm equipment, I tried to sum it up in some categories. And what you find valuable will differ between your situations, but you always have to start with your needs, your needs have to be met.
And then you have to decide what kind of data collection are you going to have? What extra are you willing to pay for? And then the practical use, how shall you manage, like William talked about, how should you manage your data? Our data is about H2S and we have published an index suggesting that we combine the number of peaks, duration, intervals and maximum peak heights into one number instead of using maximum peak as one and TVA as one and not combining them. This is a work that we have published. This first one is the proposed index and mixed model description of the exposure. The second one also have mixed model description of exposure, but it also includes the endotoxins and dust and health effects. And this third one is where we describe and evaluate an algorithm for automatic generation of index values for H2S.
The principles can be done for others too. And this is to support the use of equipment data from modern alarms, because manual handling of the alarms in the scale, it will only be a small selection. One day there, one day here.
So we have made that algorithm. More is in progress, but not yet published. So thank you for listening.
>> Thank you so much, Ase. That was great. So I think we can direct you to the Q&A as well to see what questions you got. I have one really quick question in the meantime, because I can't wait to ask it. But you collected so much data and show a very longitudinal perspective of worker feedback. And it was a lot more positive than I would have anticipated in terms of workers maybe not being bothered by as many sensors.
Obviously, if something is really heavy or uncomfortable that would kind of delay adoption. But were there some kind of growing pains in terms of when you first started collecting this data with all of these workers to maybe as you're getting data now? Have you seen kind of one element really change in terms of their kind of perception and willingness to use? >> These guys and some girls have been interested in knowing what they are exposed to since I started working with them 20 years ago. So they are an incredible group to work with actually. But of course, I see this when they don't have any docking station available, they don't do docking. So they use the sensors for alarm purposes, but I don't get any data. And I saw that partly during the pandemic, because some of them had another meetup place where there were no docking station, they only had a charging station.
And I see differences as I talked about the boss expectations. There are differences between the groups in that way. But as long as I don't identify each of them in any presentation, they're fine with me using the data for also presentations like this. >> Yeah, that's great.
And it's a really interesting point to talk about kind of how the supervisors may be communicating with workers about their intention to kind of use that data and their kind of support for it. And it kind of ties in with the idea of the health and safety management and risk management aspect of integrating all of these new sensors into the work environment. So definitely an environment really rich for research, discussion and practical application. So I really appreciate you sharing your presentation today, and I'm sure that we'll have some additional questions that we can get to during the roundtable.
So thank you so much. >> Thank you. >> And with that, we'll move on to our last presenter, Corinne Balcean. And she is the Director of Occupational Health and Hygiene for Teck Resources, which is Canada's largest integrated resource company. In her role, Corinne is responsible for leading Teck's occupational health and hygiene strategy, which includes identifying technologies to advance occupational hygiene practice. She's a certified industrial hygienist by the American Board of Industrial Hygiene and holds an MS in Occupational and Environmental Hygiene from the University of British Columbia.
Corinne is a member of the International Council on Mining and Metals Health Working Group as well. So welcome, Corinne. >> Thanks very much, Emily. And I'll just share my screen here.
Can we see that? Can you see that? >> Yes, we can. >> Okay, great. So I'll go ahead and get started. So thanks very much to Emanuele and Emily for organizing this event and for the opportunity to speak today. So what I'd like to just make very clear before we start or before I start the presentation is that really the driving force, you know, for adopting this kind of particulate monitoring technology is to provide information to our hygienists and to myself that will assist us in reducing and eliminating occupational disease. Because this technology identifies -- it helps us to identify sources of exposure and also fluctuations that occur, you know, throughout an individual shift.
And then we can then use that information about fluctuations to really pinpoint what's causing our exposures. And then so that we can limit those in their entirety. Or if we need to implement administrative controls that require changes to work processes or behaviors. And also very importantly, and I say secondarily, but it's really not secondarily, it's very important that we're really working to transition the role of our occupational hygienists from people who spend an awful lot of time generating data, you know, supporting their teams to generate that data, some cases generating that data themselves, to people who become consumers of data and problem solvers. So kind of shifting their priorities a little bit. So just by way of background, inhalation of silica is one of the greatest health and safety risks in the mining industry.
And that's because we're always moving and disturbing ground. And so, and of course, we're worried about things like silicosis, lung cancer, tuberculosis, where that disease, or in areas of the world where that disease is endemic. And as probably most people on this call would know, traditional monitoring methodologies have really not changed for about 50 years, or even longer.
And so that technology gives us, you know, one data point for an entire sample duration. So we have no information about fluctuations. And for silica, that's particularly problematic, because the method, the NIOSH method we use requires us to draw a sample for about four hours. So that's an awful long time to not have, you know, anything about fluctuation.
So Teck has been trialing real time particulate monitors in our mining work environments and also in our smelting work environment for several years now. And you know, this technology has provided us with unprecedented granular information about exposures. And so that's allowing us to better focus our control strategies, so that we're not just, you know, kind of applying blankets, we're being very, very targeted about what we're doing. However, there is still significant amount of work required before we can fully operationalize this technology in all of our work environments. And I would say that standardization of the performance criteria and usage parameters and sensor technology is really fundamental to this operationalization, because that hasn't been done yet. But it's starting, you know, we're starting that.
So I'm not going to go into a tremendous amount of detail about our trials themselves. But when I say speak about unprecedented information, and we've gone from that one line to this is a full shift for people who work in our shops, working on very large pieces of equipment. And you can see that over their shift, there's a whole -- you know, it's extremely variable. And at the same time as this kind of information, this granularity and fluctuation, our teams have done a really good job developing a cumulative exposure metric so we know the percentage of exposures that are that can be attributed to performing certain kinds of tasks.
And so I would also point out on this graph, you know, we've got a couple of really steep spikes here, we can get really excited about those spikes. However, what we found is that the spikes are not, you know, what is the greatest contribution to exposures. It's kind of those lower level wider, you know, if you look at this example of filter changing, it's kind of the area under the curve is much more significant. So that's been another one of our findings.
And so really, what, you know, our current state is we're using this technology as an investigative tool. And that's what has provided us that kind of what I refer to as unprecedented information. And so as an investigative tool, we are now -- it's mandatory that we use this in our workplace.
So as you know, we're industrial hygiene, we practice industrial hygiene, like everyone else, all of our workforce has been broken down into similarly exposed groups. And we've done all monitoring, traditional monitoring on those groups. And if we find a similarly exposed group, I'm just going to call it a SIG with an exposure over 100% of an ACGIH threshold limit value, then we have to use this technology to find out what's giving rise to those exposures. So that's our current state. However, over time, we really want to get to a place where we're using this for automation and for a decision making tool. So we'll be, you know, working on pairing and integrating this technology with other technologies so that we can have that automation resulting in a more dynamic response.
So rather than a hygienist, you know, having to send off a sample to a lab and getting a result back several weeks later, and then making decisions, we can do that, you know, kind of dynamically or in real time. And so we consider that this is a three-phased approach. We're firmly embedded in phase one, we're working on validation, and adaptation of various instruments. And then after that, we move into phase two, where we implement that technology and basically Adopt it into our workplaces.
And then phase three is what I refer to as kind of our blue sky, you know, that we don't know what we don't know yet. And so, you know, I think that anything's possible in phase three. So we're fully embedded here in phase one. And you know, it's really truly it's all we can't do this work alone. So this is all about building partnerships.
And so the mining industry has really come together in this area. And we also work with -- so Teck works with a couple other mining companies around building solutions and shared goals and objectives for the adoption of this technology. And I'm sorry, I didn't put the link here. But earlier this year, the International Council on Mining and Metals released a briefing paper on what they consider, what we as an industry consider that needs to happen for the adoption of this technology within our industry. And so it's also important that we're working with and we work with recognized authorities, researchers and influencers too, because of their leadership and leverage. So for example, we're working now with Emanuele at NIOSH, which we're thrilled about.
Because there's all -- you know, the folks that are out there leading the charge as it were, on getting this technology standardized and recognized, they need our help too. So what we bring as an industry is real world lived experience from our trials and, you know, to Ase's presentation about, you know, kind of the user experience that our workforce, you know, when they've been using these monitors in our operations. And also with our suppliers, you know, really being clear about engaging with them about the value proposition for investing in this technology to advance it. Because right now, my sense and my take is that there is, you know, very limited investment in this area. And that needs to happen in order for this technology to progress.
And I can't understate the value of working internally with your own IT expertise to get their guidance and advice along the way. Because as hygienists, we tend to speak a lot to one another rather than getting some really good sound research and development support for this work. And that's very, very important. And that's helped a lot.
And then looking at, you know, validation of the hardware and software itself, you know, does it work? Are we getting the kind of information that we're expecting to get? And can we rely on it? And then, you know, does it need to be adapted in some way? So probably, you know, one of the best examples is, we work in very cold regions of Canada, we have a mine up in Alaska. So we get down to very frigid winter temperatures minus 20, 30, or 40 Celsius. And then conversely, you know, the experience in Europe recently with extreme, you know, heat, plus 40, plus 50 temperatures. And we have metallurgical coal mining operations. So these monitors must be intrinsically safe. So we've had to work on adaptation on that area.
And there was one more that's -- oh, yes. In mining, we work 12-hour shifts. So we need to make sure batteries last for 12 hours. And then also looking at our own business processes and workflows, do they need to be adapted? And are there improvement opportunities there? And so if we can get our information in real time, and it's about particulates, and we also pair that with filters, should we be doing that analysis on site with real -- sorry, with portable FTIRs. And if so, who does that analysis? Because we do not want to overburden our hygienists. And so we're mining, we have all sorts of labs, we have metallurgical labs, we have environmental labs.
So where is the right placement for this analysis? And then how do we get that analysis or those laboratories certified? And then moving into phase two, after we've, you know, done our validation work, unequivocally supply assurance is potentially a bottleneck at this point. And so we have to work that out. And that's making sure that we support and we work with our suppliers to ensure they're in a position to provide us with robust supply assurance. And this becomes increasingly crucial as we move from right now, where we're using this as an investigative tool, to something we're then relying on for decision making. And so if I'm relying on that, I need to make sure if I order it, I can get it through the door.
And then as we deploy, making sure going back to local infrastructure considerations, because as you can imagine, every technology on every front is advancing, so do we have the infrastructure to support that. And if you look at this picture, this big long cable on the ground that this person is helping to set up, you know, we use, that's part of what provides infrastructure in mining operations. So underground operations, this is an open pit mine, so that's got electricity and that's also got connectivity. So very important we have that infrastructure. And again, with our suppliers, making sure they're aware of our deployment strategy and kind of giving them an idea of how many monitors, how many change outs we'll need, different spare parts and things like that, making sure that they're prepared to support our deployment strategy. And working with and supporting regulators and also influencing to promote adoption of this technology, they need our support to do this.
I'm a former regulator myself. And in order to get to a point where they're willing to implement this technology, recommend this technology, we need to make sure that we are supplying them with the information, the standardization, you know, of sensors and usage parameters that will give them the comfort and the assurance that if they either recommend or require the use of this technology, that we're, you know, we're not setting anyone up for failure. And then the change management process within our own operations and business, you know, anything we need to adapt and I've already mentioned things like where we analyze our samples and then training and education.
So training for our hygienists, for our workforce. You know, to Ase's point, there's a lot of work that has to be done with making sure people know what we're using, how we're using it, you know, when we use cameras, if we use cameras and all the privacy considerations there. And the people that are, we call them hygiene coordinators, so the folks that support our hygienists, making sure that they're well-equipped to use this technology. And then moving to the educational system that, you know, we are ensuring that we've got people who are coming out of universities that know how to use this technology so that it's part of curricula. And also that, you know, we're going to move into data science and at some point probably big data.
So making sure that that's also included in curriculum. Because at this point, we're still in many cases, you know, teaching traditional sampling methodologies, which is great, but we also need to make sure our hygienists have the skill set to use the technology as it evolves. And then as I mentioned, phase three is our kind of our blue sky where we expand and optimize.
And so what about wearable application software? And so that's a whole other realm that we haven't explored yet. But if we're getting all of this information, what else can we do with it? And in mining, as I mentioned, we work with a lot of very large equipment. And so if we can standardize that equipment, if it has regulatory acceptance and recommendation, then we can then influence the people who, you know, manufacture this equipment so that it rolls off of assembly lines, whether it's great big haul trucks or great big graders or dozers or pickup trucks so that it comes equipped with this technology, that the alarm systems have been kind of standardized and rationalized.
And then what do you do when there's an alarm? And also, how do you maintain the systems, you know, the cabin pressures or the air pressures or air quality? What's the scheduling for that and what's the response so that, you know, when things go back into that big shop I mentioned or showed you that graph of, our teams know exactly what they need to be doing to fix those equipment? And then area monitoring, because what I've basically been speaking about so far is all personal monitoring. So moving toward area monitoring opportunities for geofencing and perimeter demarcation and things like that. And then when we get to optimization, really, truly, this is the blue sky, you know, automation so rather than -- you know, William talked a lot about ventilation and so, you know, if we can have the sensors married up with our ventilation systems and that they, you know, we can have more or less ventilation depending on the particulate loading.
And then that it comes on and goes off as required rather than having to be a manual process. And then business intelligence, I think that's a fancy term for reporting. So can, you know, all of the data that we produce using this information, can we then combine that with other information for other reporting purposes? And again, it's still kind of blue sky here. And decision assisted tools, so rather than having, you know, getting data and hygienists having to interpret that data, that that data comes to the hygienists already interpreted and so they go, you know, into mitigation strategy. And then one step further, of course, would be artificial intelligence where data is already interpreted and artificial intelligence makes the decisions. And then, of course, advanced analytics of big data and, you know, this, you know, because one of the attributes of sensors currently that we're working with is you get a data point every five seconds in some cases.
So over time, we're going to have this voluminous amount of data and that now we've got capacity to do all sorts of advanced analytics that we couldn't even have thought of, you know, even just a very few years ago. So as I mentioned, you know, this is our journey, we're on this journey where, you know, we're very fully committed to it. We're fortunate to have a lot of leadership to be working with other mining companies, and I see some of the excellent colleagues I'm working with are on this webinar today. So we're very firmly embedded in phase one, and we have already moved into some of the initial thinking for phase two and, like I say, phase three is where we all get to kind of dream big in this area. And so if I can end with kind of a picture or more of a, I don't know, a graph, I don't know what to call this, this is my happy spot when I think of where we could be five to ten years from now.
And so this chair in this control center where all this data is centralized, that's where our happy hygienist will sit in the future is my hope. So that's the end of my presentation and I'll stop sharing my screen and turn it back over to Emily. >> Thank you so much, Corinne, that was such a great presentation. We have a lot of questions in the chat, I think, around like the day to day Things, and I just have an out of the box question that your last slide just kind of sparked for me. Which is looking at the role of the industrial hygienist, you know, throughout this presentation and the other two, we've talked so much about data science and data storage and data modernization. What do you see the role for industrial hygienists in terms of evolving and developing new skills that we didn't think they would ever need? You know, what are you looking for in terms of industrial hygienists for the future? >> Yeah, absolutely.
And you know, one of the frustrations I think many of us as hygienists have had is that we spend a lot of time, you know, our teams collecting samples, so much of that and we're people who have an absolute passion for this. We want to protect our workforce and so we get to move from people who are collecting data and spending so much time to that, to having that information at our fingertips and in real time so that we can then in real time make those decisions and we become problem solvers. And we have a seat at that table instead of, you know, those people who bother us and want us to collect, you know, data and give people letters and that -- you know, it's I think really moving us from some of the more tactical work to some of the more strategic work. And so you're absolutely right. This may, you know, be a bit intimidating or daunting.
However, I think that this is something we've been looking for for many years, many of us. So I think it's a tremendous opportunity for the field of occupational and industrial hygiene to become way more problem solvers and people who have answers at their fingertips rather than saying, okay, well, I'll get back to you when I've got, you know, my six samples for my SIG and my, you know, geometric standard deviation is close enough and I'm all good and I can say, okay, this could be a problem too. Yeah, this is what's going on and here's one of my solutions or my recommendations for how we fix this. >> Yeah, that is great. I really like your answer to that.
And it speaks to the need for all of the multidisciplinary work in this area and really the opportunity that these real time sensors are giving us to really embrace a multidisciplinary approach to occupational health and hygiene. So thank you. And with that, I&#
2022-11-16