Safely Watch and Prioritize Maintenance With AI-deep Learning and Low Cost Passive RF Wireless Vibratory Strain Sensors
Operating revenue-generating assets in harsh environments poses safety and financial risks. For example, catastrophic wind turbine blade failures are costing $1.5 billion a year to fleet owners and operators across 700,000 blades in operation globally (Wind Power Engineering, 2018), and safety incidents have increased by 22.3 percent on wind farms (Global Offshore Wind Health and Safety Organization). Sensatek discovered that passive Radio Frequency-based (RF) wireless dynamic strain sensors are a promising technology to help safely watch assets used in harsh environments, such as trending crack growth on wind turbine blades to help eliminate 3,800 blade failures occurring annually (Engineering News Record).
This technology has many applications besides wind turbines and includes airframes, gas turbines, space launch vehicles, electric machines, bridges, and natural gas pipelines, presenting a total global market opportunity of $11.6 billion (BCC Research) within the industrial, aerospace, and automotive markets. There is interest from companies such as GE, Lockheed Martin, Siemens Energy, Pratt & Whitney, and large utilities in piloting the technology.
For more information, visit here .
Transcript
00:00:00 presentation and next we have the electronics category finalist AI deep learning and lowcost passive RF wireless vibratory strain sensors representing the team today is Ramon Soo all [Applause] right hi good afternoon everyone I'm remote Soo and I'm really excited to be here all the way from Daytona Beach
00:00:34 Florida um so it's a little cold right now for me being here in New York um but I would love to just kind of share with you about you know our journey with the cotch and I like to first kind of start out with the story so I started my um career in the Marine Corps which today by the way is my birthday I am 248 years old it's the Marine Corps birthday that's like the new holiday before uh
00:00:59 veterans and also it is the birthday of citech so you know while in the Marine Corps I was in the reserves I um studied got a degree in physics at Florida A&M University while in the reserves and learned that there wasn't much I can do with a degree in physics in the Marine Corps so after I was done with my contract I uh moved down to uh Daytona
00:01:24 Beach Florida and pursued a lifelong passion the Aerospace industry and I pursued a masters in Aeronautics at Embry Riddle and it was here where I was also as my colleague uh did research on jet engines and we were working in the jet engine lab we had some GE engines in there and some uh Pratt engines and we're working on some new propulsion systems and we had parts that kept
00:01:50 failing at high temperatures and no one really knew why and so you know the Marine side of me kind of kicked back in again and I wanted to further investigate this problem wanted to be a problem solver so I went out started talking to uh the engine manufacturers like GE and um Seaman energy and I've learned that you know we weren't the only ones who are having issues with um
00:02:12 temperatures or quantifying what's going on in the most critical Parts uh such as the rotating blades and so I went out and found that utilities also use uh jet engines to generate power called aerody derivatives except when their engine basically was failing um they were forcing $3 million outages per occurrence for a new set of blades because they were having to operate the
00:02:39 engines much differently with Renewables coming online you now have more frequent stops and starts that causes the blades to uh creep but also we start we were talking with other since we're recording I won't say who it was um other folks on oil and gas platforms who are using gas turbines and compressors on LG platforms also had blades come off and it forced a $300 million outage you can see the
00:03:04 press release uh with that so I decided to uh you know we so Cent was founded to attack these problems and we used some research that came out of the University system so CCH was formed to license technology that came out of the universities to prevent these machines from exploding and by doing that we can increase safety and also Al reduce to operations and maintenance costs
00:03:33 associated with maintaining these systems and so I invented a sensor that basically uses an antenna as a wireless temperature sensor depending on the material system and a wireless strain gauge for dynamic strain monitoring so you can see here just this is just a very crude model but we our IP covers the ability to use an antenna as a sensor so for for temperature monitoring
00:04:01 the dialectric constant of the material changes linearly over the uh temperature range and so there's a relationship between the resonant frequency and a dialectric constant so by wirelessly measuring the resonant frequency we're able to know what the temperature is in real time and then on a material system that is strain dependent the sensor the antenna strain it it changes its
00:04:24 radiation pattern based on its geometry and we can get the wireless Dynamic we can get the strain reading and we over sample it to get the dynamic strain which we then convert to vibratory frequencies and so we went out back to those uh oems and the power producers and we were able to demonstrate our sensors on helicopter engines you can see the sensor on a rotating blade on a
00:04:50 stationary antenna communicating with the sensor that's spending 22,000 RPM at over 1,000° Centigrade we put our sensors on large gas turbin with Seamans and Duke energy on grid connected engines and also were used on hybrid electric propulsion systems and what we discovered was that the oems and the utilities weren't as incentivized as we had thought to really really address the
00:05:14 issue with solving the multi-million dollars and blade failures and so we had to do like any good startup does is we had to Pivot and so we then uh we're in the next era uh 35 mules program they a large wind Fleet owner and power producer and the um we had team members that said look I know that you were you know putting this on gas turban engines but we're not going to ever put this on
00:05:40 all of our engines because we're in a regulated space and you know we're just really not that incentivized to you know eliminate those failures and like the capex things and um but they were like but on the unregulated side with our the large wind Fleet you know we have issues with you know we're we're really putting the money there in the unregulated space we're responsible for the revenue loss
00:06:02 and for the cap backs associated with blade failures so we then so as the wind turban space they're growing uh you know at a rate of 8% year-over-year but for the large Fleet owners like nexera um and in venergy they're growing at a rate of 20% year-over-year but blade failures are continuing to plague the industry to the tune of $2.4 billion a year and that is the size of the GDP of Aruba and so
00:06:29 So based on our customer Discovery we learned that depending on if you're an offshore or onshore wind turbine that blade can range anywhere from a $350,000 per occurrence to a $900,000 per occurrence if it didn't take out the entire $2 to4 million machine and 7,000 of these blades are happening every year and we're going to see that problem as blades start to get larger to increase
00:06:53 efficiencies the oems are just trying to really get them out there and so we're seeing a lot more issues with blade failure and so drones are great uh but they are also prone to Misses because a lot of the cracks are barely visible and where the blade becomes catastrophic is happening is seen from the inside of the blade due to delamination and lightning
00:07:14 strikes but that's where we come in we put our antenna our our antenna sensor the size of a pizza box taking our small sensor and making it much larger to that now monitor what they call the Kill Zone within the blade and so what we're looking at is the strain which we're then converting to vibration and as cracks form is there's damage on the blades across the fleet the software
00:07:39 would then tell them how that the vibration of the blade differs across the uh entire fleet so we have four um issued patents on that technology this is the cartoon version of it so Shameless plug um I kind of already sold this thing to uh Seaman energy making them think that we already had this ready to go uh and so we don't so I would use that
00:08:03 money to actually build it and to uh sell it to them and who we have in our pipeline includes inventery from our relationship in the mhub accelerator in Chicago and next there in constellation with just their Fleet alone we can save a billion dollars per year across a fleet of 32,000 um wind turbin and that's not it you know we're working with other oil
00:08:26 and gas companies who came like Williams one Oak um and also constellation and said look can we also use your sensors to you know monitor leaks on reciprocators if we know that a combuster isn't firing properly of course in good old you know founder fashion was like absolutely we're experts at that and so they are on our road map but we're F first focused on
00:08:47 the wind turbine uh space and as we get those use cases because it seems like you know it takes a village to raise a child it also takes one to raise a business and to make a sale um so we're going to use the use case from the wind turbines to also use that in other applications like Aerospace and uh oil and gas and so we're it takes a an
00:09:10 individual to make a difference but a team to uh win a championship um but we're also a very lean we have phds in electrical engineering some of us have also came from working with the OEM before at seens and also we have participated in many energy related accelerators that helped us you know understand the voice to the customer going out to the site learning their
00:09:32 pains and gains to you know influence our road map so I'm Ramone SoDo and uh I approve this [Applause] message any questions I'd like to know if your uh sensors The Strain and temperature sensor are they combined or they sep separate yes sir great question they are um separate we did a research on 80
00:10:05 different material systems to determine what is the best uh temperature sensor that is not that doesn't change as it relates to strain and so that became our best temperature sensor and then vice versa for the strain sensor which one did not change it kept its dialectric constant relatively the same so that it's only measuring strain but it can be in the same package but they're two
00:10:27 different systems two different sensors thank you okay yes sir yeah you you touched on a very interesting point through my career which is vibration measurement um in that that sensor and mainly I see it is going on uh very high frequency devices like either it is uh turbines with air turbine or gas turbine
00:10:53 or aircraft engine now does that sensor when it's measured what does it measure frequency and amplitude or just frequency oh great great question so initially on the wind turbines what makes our sensor good compared to like accelerometers and vibration sensors is that for composits we're initially focused at the lower frequencies so anything less than 100
00:11:16 Hertz and so we're really good at that with the larger sensor size but while working also with um example here so while working with like Pratt for instance and other OEM they want you know the higher frequency stuff and so with that we're we're being compared to our tip timing so we put multiple antennas in different locations
00:11:47 so that we can get the frequency and the mode at that particular time so it's like they're asking hey do I if I get the frequency of the blade at the higher frequencies CU it changes its frequency from zero the 360 and so the question is where at in that you know chart did you have that frequency and so that's what we're working collaborating with the OEM on and so right now the higher frequency
00:12:13 stuff we're aiming for 100,000 Herz which is wait 100 khz um we're not there yet but it is on our road map but the wind turbine you know family they were like we want to buy this right now and the oems were like hey we just want to kind of study this to St um and then we'll think about using it so it's on our road map but it's not a key priority just yet but
00:12:36 good question thank you thanks for the presentation you mentioned that it's a size of a pizza box but it's not as thick as a pizza box I'm thinking about aerodynamics of the wind turbine is there any issues there right great question so um our sensors vary in size depending on the uh application so we've made really small ones the size of
00:13:01 a postage stamp to a ladybug and we've also made some the size of a sheet of paper but the thickness is less than .1 mm in thickness and so it's similar to what you would use like a capton you know mask or a Rogers material and the sensors are mounted initially on the inside of the blade for blade root delamination so we're essentially catching what the drones aren't catching
00:13:25 um we are thinking about you know applying it on the outside of the blade like closer to the tip but most of the insurance claims are coming from the Kill Zone and the blade route so it seems like that is probably the best place to to start on the inside of the play one last question sorry to keep you stood up that what's the Frequency with which blade
00:13:50 failures actually occur that is a million dooll question so on a wind turbine since we B recorded on a wind turbine maybe the natural frequency is around 8 to 9 Hertz so we're anticipating that as the blade starts to crack or as it starts to delaminate that it's going to lower the frequency but just to make sure that we catch other failure modes that we're not
00:14:15 aware of because the O the customers the fleet owners They Don't Really there's an educational piece where we have to let kind of you know help lead them on that but we're capturing between zero to 150 Hertz so that we can I compare what they're saying on the Drone data and on the um human-based inspections to what we're also seeing on the vibrations so with that pilot with Seamans we're
00:14:41 hoping to better understand that correlation and how the um the failure modes how it changes the vibratory frequency can I ask a quick follow up please how frequent how often do blade failes occur with one customer that we're talking talking to right now they had 40 last year so it was about $16 million uh Bill and they're anticipating more to happen but also it's not just
00:15:09 the blade failur when they identified a category three or category 4 crack they're now having to do six drone inspections instead of only doing one drone inspection and it's now increasing the operations and maintenance costs on the uh project and they have to lock up that amount of money for like 5 years based on how they financed the project so where we come in not only are we
00:15:31 catching blade failures but it's also eliminating the amount of drone inspections that have to be done and when you do a drone inspection you got to shut the machine down you got to coordinate that there's more eyeballs on it there's more om andm involved on it and they're losing Revenue when it's down doing an inspection and so there it's twofold reducing the onm and also
00:15:52 mitigating uh blade rout uh failure One customer had 40 not all customers but one had 40 uh blade failures the average Fleet is about 10,000 turbines thank you thank you thank you