A safety vest that alerts a roadside worker of oncoming vehicles is in the works. Hear about the InZoneAlert safety vest from Virginia Tech’s Tom Martin.
In development now, the safety vest senses the accident avoidance signal from modern cars and trucks. The vest is equipped with a radio receiver which triggers flashing lights and audible alarms when a vehicle is too close.
Professor Martin, of the Virginia Tech Bradley Dept. of Electrical & Computer Engineering, explains the technology involved and the roadblocks to manufacturing and sale.
TRANSCRIPT
(:00)
Brandon Nys: Welcome to Safety Experts Talk. Visit our website at CreativeSafetySupply.com/podcast.
(intro music)
(:14)
Prof. Tom Martin: We were looking at three conditions: A car that’s passing by in adjacent lane, a car that’s a half a lane width off, and car that’s heading right at you.
Dan Clark: If a vehicle is about to collide with worker, wouldn’t it be great if the worker would somehow get a warning? Even just a few seconds’ warning could save a life. That wish is coming true with the InZoneAlert safety vest.
Hi there, I’m Dan Clark. Today, we’re with Professor Tom Martin of Virginia Tech. Hello professor.
(:44)
Prof. Martin: Hello, Dan. How are you doing?
Dan: I’m alright. The InZoneAlert safety vest. Tell us about that.
Prof. Martin: The InZoneAlert safety vest is a collaboration between the Electrical and Computer Engineering Department and the Virginia Tech Transportation Institute, VTTI. They have a larger research effort in the connected vehicles initiative. And the idea is, in the near future, cars will have GPS units and short range radios and allow the cars to talk to each other to do collision avoidance and things like that. The idea that my colleagues and I had was to take that same technology and see if it was feasible to give a worker at a roadside construction site a few seconds of warning about a car that was heading toward them.
(1:25)
Dan: So that would be any vehicle that has this new technology.
Prof. Martin: Yes. We can’t do this right now ‘cause these cars aren’t widely on the road but my colleagues down at the Transportation Institute tell me it’ll just be a couple years before they expect to see these cars widely available. The idea is any passing car that has one of these systems on it — and in the near future we hope nearly every car out there will these systems — we’d be able to communicate with it and give the worker a few seconds warning as well as give an alert to the driver to let them know that they’re driving toward a worker on foot.
(1:55)
Dan: So this is just a typical orange vest.
Prof. Martin: Right.
Dan: It has a device that’s a receiver in it and some kind of warning system?
Prof. Martin: Yeah. So, I should say that we’re just at the prototype stage at this point. Last year we established the feasibility of low false alarms. And that looks feasible. It looks like with about 90 percent accuracy we can give a worker on foot about five, six seconds of warning for a car that’s coming at them at 35 miles an hour. So once we established that feasibility what we’re looking at right now is: What’s the best way to provide the alert to the worker?
(2:27)
My students right now are taking off-the-shelf construction vests — orange vests — and we’re looking at flashing lights, warning alarms as well as some vibrator motors to give them tactile feedback. We’re undergoing testing right now to see which combination of those seems to be most effective.
Dan: Sounds like the guy would look like a one-man band with all of these devices all over. Is it fairly weighted?
Prof. Martin: It is now because we’re just prototyping things. At this stage of process we’re just trying to see what works.
(2:56)
Dan: Mm-hm.
Prof. Martin: But the idea is, once we know what the right alerting mechanism is, it should be no problem shrinking this down to something you won’t even notice. And the reason we picked the construction vest is because we know the workers will already be wearing that. We don’t want them to have to remember to wear something else in the morning when they go to work (or in the evening if they’re on a nighttime jobsite). We want to fit into their existing routine and we know that the workers will be wearing a vest.
Dan: Now you have a closed worksite where, what do you have … mannequins out there next to the vehicles as they go by?
(3:28)
Prof. Martin: Right now what we’re doing is we’re testing in the lab. So, we’re inside a building with a surround sound speaker system. When we did the feasibility testing just to see if the GPS units were accurate enough to give us the kind of warning we wanted, the Transportation Institute has a closed smart road. They use it for a wide number of studies. So, they can create rain, snow and other driving conditions. So we had this closed road that we were using to test the GPS systems with.
Dan: So, for me, GPS doesn’t seem that accurate. With this new technology that you have is GPS granularly accurate enough to inches where, I suppose, it would need to be?
(4:05)
Prof. Martin: Well, not down to the inches but what we were looking at was three conditions:
— A car that’s passing by in adjacent lane;
— A car that’s a half a lane width off;
— A car that’s heading right at you; and
Dan: Uh-huh.
Prof. Martin: And what we wanted to say was:
— We will give you an alert if the car’s heading right at you;
— We’ll give you an alert if the car is a half a lane away and you should pay attention; and
— If the car’s a full lane away — well we shouldn’t give you any false alarms, then. That’s what we assume will be the normal distance.
Dan: Mm-hm.
Prof. Martin: With 90 percent accuracy we can give the proper amount warning of about five to six seconds at 35 miles an hour.
(4:39)
Dan: Wow. Well, I know a lot of work zones I’ve seen where the workers — especially in a paving project, for example — they’re right on the edge of the lane of traffic. So you would have warnings happening every moment.
Prof. Martin: That’s the situation we’re trying to avoid. We think it’s accurate enough that we will be able to avoid that. You have to take the width of the car into account too, because the GPS units will be mounted, we presume, somewhere near the middle of vehicle. So, we understand that condition might happen but we think we’ll be able to avoid too many false alarms.
(5:09)
Dan: What if the worker’s under a tree or in a tunnel? Does the GPS work there?
Prof. Martin: Probably not. The best I can say is if we can provide some workers with a warning, that’s great. I can’t guarantee this will work under every condition.
Dan: Yeah.
Prof. Martin: If you can’t get the GPS signal, you can’t get the GPS signal. There’s nothing we do about that.
Dan: They wouldn’t have any warning without the vest anyway so it’s not like they’re losing anything by putting it on.
Prof. Martin: Right.
Dan: I know that you have, as a professor, a lot of stake in this but you also have students. Whose idea was this?
(5:43)
Prof. Martin: It actually goes back a long way. A number of years ago I ran a class with a colleague in industrial design. What people now call Internet of Things [IoT] in the research community goes by the name pervasive computing. I think this was 2007. We were looking at “What are the applications of pervasive computing in construction sites?” In particular, what I’m interested in is “How can I give an individualized warning to a particular construction worker who’s at risk?”
(6:10)
One of the scenarios that came up was roadside construction sites. The idea the time, if I remember correctly, the students had an idea for some smart barrels where the barrels would actually give the warning.
Dan: Hm.
Prof. Martin: Some time went by. Some colleagues of mine at VTTI found out about the class and then we wrote proposal to look at not roadside construction sites but actually closed construction sites and construction vehicles rather than cars passing by.
Dan: Mm-hm.
Prof. Martin: So we wrote a couple proposals on that and then this connected vehicle consortium came up. That’s when we decided to look at the situation of passing cars.
Dan: Ahh! OK.
(6:44)
Prof. Martin: Yeah and one of the neat things to me about this is — I’ve been working on wearable computing for over 20 years now. I’ve been building wearable computers since the early 90s — this is one of few cases where we could actually build the wearable today if somebody would give me the engineering money to shrink things down.
Dan: Mm-hm.
Prof. Martin: It would be no problem to build ___ for the GPS unit and the radio. What’s actually going hold things back is the deployment of the vehicles — having these collision avoidance systems.
Dan: Right.
(7:11)
Prof. Martin: Until those cars are on the road, the system as it is right now won’t work. But the technology’s there. We could shrink the GPS units down, we can shrink the radios down. This would be really not much of a problem. There’s no magic bullet that has to be found. It’s just a matter of throwing engineering time and money at it.
Dan: Well, while you’re waiting for these vehicles, could you have a triangulated radar where it automatically monitors every vehicle?
Prof. Martin: It might not be a radar. I think there are things that you could do like that along the construction site. I think that would be totally possible.
(7:44)
Dan: Ahh, OK. But the vest would still be the warning device for the worker.
Prof. Martin: Yes. All we need is some way to get the trajectory of the vehicle who’s coming toward the construction site …
Dan: Right.
Prof. Martin: … and have that transmitted to the vest.
Dan: How long do you think it’s going to be before you get the information and decide whether it’s marketable?
Prof. Martin: Given the interest that I’ve seen in the vest — I’ve actually been surprised at how much interest that people have shown in the vest — I think it won’t be very long. Again, I think the real sticking point is going to be “How long is it ’til we have enough cars on the road with these collision avoidance systems?”
(8:21)
Dan: Mm-hm.
Prof. Martin: That’ll be the real hold up.
Dan: Well, I can tell you why I think — if you’re surprised by it — because I know that struck-by fatalities in work zones have been 600 per year in the United States. When you look at those statistics, you’ve got to say “Well, there’s got to be some solution out there” and this sounds like a really good one.
Prof. Martin: Right. A colleague of mine was talking with somebody who owns a construction company. What the owner said was the best he can do right now is have some of his workers go down to the nearest bar and offer to give people rides home. That’s the best thing he can do to provide safety to his workers at this point.
Dan: (laughs).
(8:55)
Prof. Martin: And he was … sounds like he was only half joking when he said that. The impression I got was he’s actually done this.
Dan: Well, I wish you the best of luck. And when do you think you’ll be done with testing?
Prof. Martin: It’s an iterative process. We’ve already made a second version of the vest. I had the students make an initial version of the actual vest with the alerting mechanisms in it back in the class that I taught in the spring. I have a graduate student who’s working on a new version of it right now and once we start looking at the initial results we’ll probably make a third version of it and keep refining it.
(9:26)
Then, what I’m hoping to do, is go around and find some more funding sources to help develop this further and try to make something that’s really a finished prototype that we could take out to a manufacturing facility and have actually fabricated.
Dan: Mm-hm.
Prof. Martin: Say, 10 or 20 units.
Dan: Looking farther down the line do you plan to license it to all vest manufacturers or would you just go “well, we’re going to have one company make ‘em and they’ll have the exclusive.”
Prof. Martin: I’m not sure that’s entirely up to me. If it were up to me, then everybody would be able to make these things.
(9:57)
Dan: Mm-hm. Well, however it can get out to the workers to wear it, the better.
Prof. Martin: Yes, I agree.
Dan: I know that improving a safety vest is not your first construction zone safety invention. In 2013 you came up with the — what is it, a carbon monoxide monitor for hard hats?
Prof. Martin: Yes.
Dan: Tell us about that.
Prof. Martin: Well, like I said, we had this class a few years ago and an indirect outcome of that class was we came across somebody who studied carbon monoxide poisoning in construction sites. Right now the best you can do is have an environmental sensor for carbon oxide to get people a warning that the environment has too much carbon oxide in it. But it turns out different people absorb carbon monoxide at different rates.
(10:40)
Dan: Mm-hm.
Prof. Martin: So the same environmental level will have a different impact on people in the location. Particularly if somebody’s a smoker or anemic and other things they will succumb more quickly to carbon monoxide poisoning.
Dan: Yeah.
Prof. Martin: You know the … if you’re in the hospital they put the blood oxygen sensor on your fingertip?
Dan: Sure.
Prof. Martin: There’s a similar technology out there for sensing the blood levels of carbon monoxide. It just involves some different wavelengths of light. The idea was if we could sense the blood levels of carbon monoxide then we could give a worker and individualized warning whether they are about the succumb to carbon monoxide poisoning.
(11:14)
Dan: Wow.
Prof. Martin: We realized for a construction worker you can’t put something on the fingertip. They need their hands free. Then it became a process of trying to find someplace on the body where we could put this, where it would work reliably and we had a reasonable guarantee that a worker would have it with them at all times. And again — the same ideas with the vest for the roadside construction sites — the idea was we could put this sensor in the hard hat, which we knew the construction workers would be wearing. And that actually turns out to be a pretty good location to put it, because instead of passing the light through the skin like happens with a fingertip, we can actually send it through the skin, bounce it off the skull and catch it when it reflects back out. It’s a different type sensor but it works the same way.
(11:54)
Dan: So the internal headband inside the hard hat is against the forehead. Is that the sensor?
Prof. Martin: That’s where the sensor is, yes.
Dan: Wow!
Prof. Martin: And the thing is we didn’t want to test it by actually exposing people to carbon monoxide. It works on the same physics principles as the oxygen sensors. So, what we really did was, we just used a blood oxygen sensor that works in that mechanism. That’s called a reflective sensor. So we put a reflective blood oxygen sensor in a hard hat — in the strap — and used that to measure people’s blood oxygen levels. Then what we had them do was perform some tasks, because if you’ve ever worn one of those blood oxygen, you realize just moving your hands around — just moving the sensor around a little bit — throws the readings off quite a bit.
(12:35)
Dan: Oh.
Prof. Martin: Because, it turns out, it’s really measuring the amount of blood that’s flowing through and so small movements of the body actually change the amount of blood flow in localized areas.
Dan: Mm-hm.
Prof. Martin: So, what we were looking at is how much we guarantee could we provide somebody to give them a warning within a few minutes of entering a location. Unless there’s really high levels of carbon monoxide, you won’t pass out instantly. It will take you several minutes to succumb to it.
Dan: Right.
(13:00)
Prof. Martin: So we had done some calculations to say, at the environmental levels most people were concerned about, it would take a normal person about 12 to 15 minutes to be at risk. And then, based on calculations we’d done, we were trying to see if we could give a warning within — its been a while now but I think it is either five or six minutes.
Dan: Mm-hm.
Prof. Martin: And we found that with more than 99 percent confidence we think we can get somebody a warning within five minutes …
Dan: Wow.
Prof. Martin: … even if they’re moving around. So, we had people like hammering on things and walking around and sweeping. Probably the only thing that we wouldn’t be able to guard against was if somebody was running a jackhammer.
(13:35)
Dan: Oh.
Prof. Martin: If you’re doing something where you’re really bouncing up and down and you’re doing it continuously for several minutes, we probably wouldn’t be able to give you a warning ‘cause then the hat would just be in constant motion.
Dan: Mm-hm.
Prof. Martin: But just walking around, sweeping, using a hammer, other things, you’ll stand still just long enough for us to get a reading.
Dan: Well, did anybody pick up on it? Did NIOSH say “Hey, this looks like a good one. Manufacturers take this …
Prof. Martin: No.
Dan: … take this idea.”
Prof. Martin: No. Haven’t had anybody pick up on that one.
Dan: Hopefully somebody will take that on, but …
Prof. Martin: Right.
(14:06)
Dan: I think you’ve got a cottage industry there at Virginia Tech. You could be making a lot of safety products.
Prof. Martin: Well, like I said, my background is actually not in construction. My background’s in wearable and pervasive computing. But the idea is the intersection of safety concerns and this notion of the Internet of Things and pervasive computing is we can give much better individualized warnings that we could with environmental sorts of sensing.
Dan: Well, I’m worried about wearable computing systems for going to Las Vegas. Come on, let me go gamble with your systems! Can I do that?
(14:41)
Prof. Martin: You can’t with mine, though I’ve read a story of one of the very early wearable computers was to do exactly that. I think it was card counting.
Dan: Yeah.
Prof. Martin: Back in the late 60s.
Dan: Yeah.
Prof. Martin: But, no. I haven’t done that. I like my legs.
Dan: Yeah, no kidding!
Prof. Martin: (laughs)
Dan: Well, what other kind of wearable computing projects have you worked on?
Prof. Martin: So, the non-construction things I’ve worked on …one of the things I’ve been doing is we’re making electronic textiles. This is sponsored by the National Science Foundation. So, we’re actually weaving wires and other electronics into the fabric on a loom with normal yarns. So, we’ve done a number of systems based on electronic textiles.
(15:21)
Dan: And what’s the purpose of that?
Prof. Martin: It depends on the application. So, right now we’re working on a garment that can classify what your activities are while you’re wearing some sort of physiological monitor like a heart rate monitor or a blood pressure cuff. It will automatically create a diary those activities so that you can look at the physiological data and annotate it with what your activities were.
Right now, for instance, if you have a heart arrhythmia, it probably won’t show up in the doctor’s office. What the doctor will do instead is give you what’s called a Holter monitor, which is a wearable ECG unit, and ask you to wear it for 48 or 72 hours. And you’re supposed to keep a diary what you’re doing while you’re wearing it.
(15:58)
Dan: Oh!
Prof. Martin: And people are really bad at doing that.
Dan: Yeah, no doubt.
Prof. Martin: So, the idea is we can give them the clothing that they can wear at the same time that will automatically classify the activities so they don’t have to remember to do it.
Dan: That’s even better than the medical bay on Star Trek where they lay on the table and it reads it. But you get to walk around and wear the clothing.
Prof. Martin: Yes. Everything we’ve done is … you can wear it around. I’m an engineer so I make really ugly clothes. This current project is actually a joint project with the Apparel Design program at the University of Minnesota. So, hopefully, we’ll get better looking clothes out of it than what I typically make.
(16:30)
Dan: A high-tech runway show in New York. That’s what I’m looking forward to.
Prof. Martin: Yes, that would be great.
Dan: Well, this has been fantastic and I hope you have really good luck. And I want to check in later when you’re done with your testing on this special vest. I really appreciate your time talking with us and good luck with it.
Prof. Martin: Thank you Dan. It’s been great talking with you.
Dan: That’s the InZoneAlert safety vest. Our guest is Professor Tom Martin of Virginia Tech and the Bradley Department of Electrical and Computer Engineering and Safety Technology. I’m Dan Clark.
(17:06)
(outro music with voice over)
Brandon: Thank you for joining us on Safety Experts Talk. If you have suggestions for future podcasts, send them to podcast@creativesafetysupply.com. For more safety experts talking about safety news, OSHA regulations, PPE, lean, 5S, or Continuous Improvement, go to CreativeSafetySupply.com/podcast.
(17:38)
END
Virginia Tech video:
http://vimeo.com/137485568
Tom Martin, Kirsten Hines and Virginia Smart Road photos © ℗ 2015 Virginia Polytechnic Institute and State University; Road workers in orange vests provided by The National Transportation Safety Board