okay so to introduce our speaker for today marisol barrero is a certified professional ergonomist and has worked for toyota north america in their production engineering division safety group for 15 years she currently manages development and integration of safety related technology across their manufacturing operations she also externally serves on the boards of the automotive exoskeleton group as well as the astm exoskeleton center of excellence marisol is a regular speaker at various safety technology focused events we're very excited today to have marisol speak to us regarding emerging technologies to enhance safety at toyota motor north america welcome marisol thank you can you hear me okay i can okay awesome always great to confirm so uh good morning good afternoon good evening everybody depending where you are so i'm really excited to be here today and talk to you a little bit about what we're using at toyota and really just an overall disclaimer to this is again i'm not pushing any of these technologies on to you i'm simply simply explaining what we're looking at what we're finding to be of value and how we're proceeding within our organization some of these may apply to you some of these may not um again it's really just what we're finding uh to be useful my intent here is just you know hopefully this can shed light on some of these technologies and um and and certainly we can progress from there uh as an overall safety uh entity advance my slide here okay so um so jessica did a great job of introducing me so my background is ergonomics and uh really about 13 years that i've been at toyota i've been implementing ergonomics and in some sort of capacity i've always been within the the safety organization um everything from you know new model uh development so making sure that our vehicles meet the standards for ergonomics for production to keep team members safe uh to you know auditing program development um basically almost all areas of ergonomics uh a couple years ago up until a couple years ago i was the regional ergonomics manager so for all our plants and headquarters across north america so again my my background is framed in that i've been lucky enough being in safety for a number of years to really take in in addition to being a cpe take in a lot of other aspects of safety outside of ergonomics as well getting that exposure to industrial hiding hygiene machine and robot safety you know really the full gamut of safety operations so starting out with that i just wanted to give you a super quick glimpse into organization when i say toyota motor north america i'm primarily talking about the manufacturing side of the operations uh we're slowly working more and more with the sales and marketing and financial side of things but it's it's mostly the manufacturing side that i'll be talking about today so basically i'm within the safety group of toyota and essentially it's a corporate safety function as some of you might know it and we support all our manufacturing plants so we have 15 across north america we've listed them there the majority in the u.s we do have some in canada and some in mexico as well so we support them in all aspects like i noted before everything from lockout to you know managing their kpis around safety incidents et cetera full gamut our headquarters our operations headquarters is located in plano texas r d is in ann arbor michigan and r d handles some of the the ergonomics for the design of the vehicle so for example the cockpit design etc seat design and such so we don't dabble in that here in production engineering where i'm at and then a good segue into engineering location which is georgetown kentucky where i'm personally based so just before we get started just want to quickly share our corporate safety statement and policy which is toyota cares which essentially cares as an acronym that stands for compliance accountability reporting excellence and sustaining so this really guides a lot of our activity we have a goal of zero injuries and really a lot of what i'm doing now with regards to innovations uh and i also have communications as a as another responsibility but a lot of what we're doing now with regard to safety innovations is we're really grounded in that e for excellence which is uh excellence in safety through innovation and continuous improvement and this is i think a good way to introduce a lot of these different things that i'm going to talk about some of them might seem a little far off into the future in terms of application but the way i really view it things is is with this quote by arthur c clark and space odyssey uh responsible for that so the only way of discovering the limits of the possible is to venture a little way past them into the impossible so you know we're testing uh things as we go along seeing as they fit seeing if they don't fit seeing how we can improve things so that they do fit so so just really keep that in mind and keep an open mind with all of these things this is a safety technology roadmap that i've had in place the last few years when i i was i was still the ergonomics regional manager when i started dabbling in this and i started looking at exoskeletons primarily and that really paves the way to this role that i have now so it's really interesting a lot of this stuff is new to everybody including to me so it's a very interesting process trying to put this in place as compared to the more established traditional ergonomics discipline so um so these are some of the things that we wanted to hit in a linear fashion you know throughout the last few years some of the things that we started working on in 2018 we're a little bit more reactive based on some incidents and some issues that we have had with some of our equipment you can see some of it is about data management and cranes and powered industrial vehicles and such and then we started getting a little bit more proactive so particularly in the area of exoskeletons so trying to see trying to explore those technologies and then paving the way for other activities so exoskeletons and suits and other body regions outside of the shoulder computer vision wearable sensors virtual reality augmented reality experiences and then all the way through artificial intelligence and intelligent plant concepts we are currently compiling a technology roadmap for 2021 and beyond so the next few years and that that's not ready yet um but uh it's it's it's interesting how it's developing and it'll be it'll look a little bit more complex than this one which is very linear in fashion so i wanted to hit some of these items that we're working on today and my agenda for today to share is uh it's going to be a lot about exoskeletons because that's where we started and that's a program that we do have implemented but i also want to share some activity that we have going on with wearable sensors computer vision vr and ai and again as i mentioned before any mention of specific commercial products or services it's not a direct endorsement by by toyota and certainly not by niosh either so first exoskeletons so some of you may be aware of exoskeletons i know there's a growing body of evidence for exoskeletons out there so with regards to shoulder exoskeletons we currently have more than 500 devices in place across six of our north american vehicle assembly plants in the vehicle assembly plants that's really where we get the overhead work so we're these are difficult operations um where you can see some photos there they're they're either installing parts uh or fastening the various parts and and such on the vehicle uh as part of their regular work and it's been implemented as ppe and multiple applications so uh paint body weld and assembly shops across those six plants with the overhead work we do have some engine plants uh and some unit plants across north america but again those do not have the overhead work so uh for us right now it it is required ppe for a process i'll expand on this image right here a little bit later on but when do we decide when it is pp so for that particular process so first the process is flagged by a selection tool uh which is essentially and i'll cover this in a bit but it's a mathematical calculation based on the percentage of the cycle time with overhead work postures so then as the process is flagged we do some more analysis so obviously the process has been flagged because there's overhead work so by overhead work the hand is above the shoulder for some portion of the job and we start looking at different controls that may possibly be feasible other than exoskeleton so we we don't want to jump into you know putting an exoskeleton on on anybody that's not necessary but we start assessing different possible controls but if no other control is feasible in the short term then we start to consider the exoskeleton more closely and then we examine to see if the exoskeleton is going to pose any sort of additional safety risk to the process so an example of a risk that an exoskeleton might pose is you know just entanglement uh with something or it could be something outside of safety it could be for example it's somehow going to activate the light curtain for the equipment so the process is going to stop or something like that so we make sure that the exoskeleton is a good fit and it doesn't pose any additional risk and then at that point it's required for the process and that means for all the team members working the process so what we've seen with regards to shoulder exoskeleton impact um so you know we we didn't just start this yesterday we started this in in 2016 so it's been quite some time exploring this and it really started as a grassroots effort so there was no top down hey look at exo skeletons or implement this there were no executives really guiding what we were doing but it was just uh hey this is something that could really be useful for team members and some of these operations where we really can't do anything else so we you know we we can't rotate the vehicle on its side uh so what what we saw when we started uh trialing some of the products the shoulder exoskeletons as they were the most robust uh devices at the time so the the ones for the other body regions weren't really as strong so that's why we started with shoulders we started seeing really reports of greater comfort using the devices as compared to not and lower ratings of perceived exertion or rpe a term that some of your ergonomists out there might be familiar with and we actually did do a couple of emg studies or electromyography study to look at muscle fatigue which i'll talk about in a little bit but our progression with shoulder axle skeletons this is really where the uh the vas experience is in and we started out again just being aware of the devices in 2016 and the plan started dabbling uh with trialing the devices in 2017. we developed a lot of our system so we wrote our regional standard uh noting that it's required pp where it's identified and safe to use in 2018 um and then really we've been working on um just just honing the the overall standard so in 2019 it became a requirement at the end of 2019 and in 2020 and beyond by maintenance i mean we're really just trying to maintain the standard that is out there and work with the plants to make sure that it's actually being followed and helping them so i want to share a little bit of the electromyography study uh that really paved the way for us to see that these devices would really be useful with overhead work so in 2017 we partnered with iowa state university dr gillette and terry butler from lean steps consulting to complete this emg study the first one looking at specifically the levitate airframe so in case you're wondering in our operations right now for the shoulder we use a mix of the levitate airframe and the autobot paxo uh shoulder exoskeleton so we've uh we've looked at other products that we've um we've we've kind of stuck with those two for a variety of reasons here you see an image of i think this is dr gillette the angle is kind of odd but uh he's uh he's establishing the maximum voluntary contraction for one of our team members at canada plant what i refer to as tmmc so sorry if i go back in my lingo so i'll i'll call it canada plant for all of you uh essentially that's what an emg study aims to do so establish that maximum voluntary contraction for that muscle in this case on the right hand side he's doing that for the bicep and then we we put exoskeletons on the team members while they do their process and then we look at their um their effort as compared to that maximum effort and study that more closely with them without the exoskeleton so in in canada so we looked at 10 underbody processes that were totally different in terms of the amount of time that was actually spent overhead and again they're very very different processes the only commonality was that they were doing some sort of overhead work so across the board and again the muscles that were examined were the deltoid um the biceps and the erector spinae so in these 10 processes overall the study found a significant reduction in the percentage of mvic for the deltoid and the back muscles when wearing the levitate airframe you can see some numbers there the bicep was studied and it was a little bit lower but it wasn't statistically significant now it's i think it's more interesting when you look at the various processes in a little bit more detail so we we had again that figure i gave you earlier was average across the board but the percentage mvic was as high or as as 16 for the deltoid in this test that was called the left-hand absorber job test however if you look at psa uh it actually showed an increase in mbac for one for for that particular process so um in this case for psa uh the effort the mvic was actually higher uh for the deltoid when wearing the exoskeleton uh of course it didn't show that association for the back but the the key point here that we started seeing is that you know we can't really look across the board how we started and we said hey we're going to put shoulder exoskeletons on um on all the problems of the overhead work it really depends depends on the specifics of the process even then even though it is overhead work so the effectiveness varies by process which really led us to craft a tool to numerically tell us when to implement so our process for determination so there's a second emg study that we did with iowa state and this one was lab based and i'm sure they can explain it a lot better than i can but essentially we will utilize the acgih upper limb localized fatigue threshold limit value which of course is an external tool out there from acgih and um we we also utilize our internal ergonomics assessment tool which is a proprietary tool called tiva to go to ergonomic burden assessment that looks at courses postures uh attack time you know the cycle time for the overall process weights etc uh and it really breaks up the process into all these different details so um so again they looked at mvc and they mapped out the various postures so those little posture images that you see there are directly from our tool so in the tool you look at a process and you you note the number of seconds that the team member is actually in all those different posture codes so we were able to map out uh you know how this tool actually mapped out to our tiba uh and what what this basically means so an example there so posture code eight so if you look at um i'm not sure if you can uh see my little arrow there if you look at posture code eight uh so if your the duration fatigue limit is 7.5 seconds for a 60 seconds hack time so if we have a any process that um is above 7.5 seconds and it's a 60 second tech time this process would be flagged and again it's just an initial flag we would need to look a little further to make sure that an exoskeleton is appropriate so this is what our tool looks like it's uh it's very simple it's a this is information that is already available from the ergonomics assessment you can sort of filter out from the existing tool so we're looking at the various posture codes and seeing if any of them exceed that number that has been signaled out by the acgih upper limb curve and then this list that you see here is an actual list of filtered processes from tmmc or canada plant or north assembly shop in the chassis section where they do have the overhead work and this is a list of processes where oops sorry i advanced there that has been filtered so essentially uh this these processes have been flagged for possible exoskeleton implementation so then i've explained these two items on top so again we use that tlv and we use the our internal screening tool and then as i explained earlier we examined there any other feasible controls to put in place and we do a safety assessment um and if all of this passes then the shoulder exoskeleton is required for everybody in the operation uh this this may sound very far-reaching but it's it's really a very small application so uh you know when we go through the board of all the different processes we're probably talking about you know less than two percent of the processes in a plant uh or even maybe less than one percent these plants are huge would really require an exoskeleton as pp so um if you walk into one of our plants it's not something that would stick out at all so i know there are lots of forums and debates as to whether exoskeletons are ppe or if they're tools or they're some sort of engineering control as yet undefined we wanted to call it something uh without really concern for what you know is appropriate for others and what it may be called in the future so we really wanted to put something in place for something that works for us for toyota we labeled it as pp for the shoulder exoskeletons because using the hierarchy of control the risk is still present the exoskeleton protects the team member from exposure and really calling it pp really allows us to focus on what we really want to do so visualize the need for stronger controls prior to assigning any exoskeleton for us at least it really allowed quick and easy alignment with existing ppe policies so we actually took the the internal respirator standard that we have in place and actually adapted it for exoskeleton so it was a really good foundation to already have and the other item is that it allows us to have risk based decisions so things are not static um you know they're ever changing honestly they're a little too dynamic so ppu requirement is based on the process time and overhead work so that means when a process changes and it may change every so often the requirement for ppa may change so resulting process may may require so a new process may now require or it may cease to require an exoskeleton so um in case you're interested so i know i've gotten the question before regarding what do we do of our process changes and all of a sudden exoskeleton is not required uh what we do is essentially as long as assuming there is some overhead work in the process we actually give the choice to the team members and particularly our indiana plan they've reported that when they when they've seen such changes about 50 percent of the team members choose to retain their device even though it's not technically a requirement so it's it's interesting um i think they're certainly seeing some some value it's a good segue to to note the impact so indiana really have to uh praise them they've they've done probably the best job of implementing the exoskeleton standard and being uh you know very consistent and what they've actually achieved is very commendable so they've actually had zero shoulder injuries in the impacted processes since deployment of the unit so um they they technically made a requirement at the end of 2019 but they had team members here you see some in pain shop wearing them as early as 2017 2018 uh starting at 2018 they actually saw really zero shoulder injuries with in those processes where they were designated as required ppe so really kudos to them for showing that and again this is really just icing on the cake we've seen a lot of other positive benefits with you know better team member feedback comfort et cetera from the devices so definitely we are not perfect in terms of implementing the exoskeletons in any way shape or form we have a lot of challenges like anything else we did look last year and throughout this year to understand the the causes or the reasons for no implementation even though uh the processes have basically been signaled as as required uh exoskeleton process and the top four reasons are noted there so number one reason was dynamic work i'll talk about what this means um then number two was bending and calves so the exoskeleton was in compatible space constraints and height of reach some of these devices basically don't don't support your arms beyond a certain angle so if you're completely overhead with your arms out stretched above your head you may not actually get that benefit so uh dynamic work what does that mean so dynamic work is that the the team members are spending that cumulative time overhead but they're going up and down a lot so they're you know hopefully you can still see me on camera so they're basically installing something they're coming back down grabbing something going back up so that up and down was really uncomfortable for some of the team members because these are uh passive shoulder devices and they're not you still feel that a little bit of resistance when you're coming down so um so that sense made it a little bit uncomfortable so we're we're trying to uh counter-measure that through more on-site plant support of course we haven't really been able to travel this year which has been really difficult and we're trying to understand if some exoskeletons are more applicable with certain concerns as compared to others and we're also trying to do a study with the university of waterloo in ontario to understand what is dynamic how do we define dynamic so that we can put these devices in the processes where they're going to have the most benefit and they're also going to be comfortable and beneficial for team members so we're just trying to hone the tool that we have in place just through again continuous improvement of our operations so now i want to share a little bit beyond the the shoulder i want to talk about the hands and some of the items we've considered there so of course we have repetitive short cycle time jobs we have msds and really you know a lot of the areas of the body so upper extremity uh being one of them hands uh back et cetera we did a trial of the biocervo iron hand which is an active powered hand exoskeleton it can be worn alone or under the required glove so there you have the actual bioservo glove and that that upper photo there that is from bioservo and the glove assist while performing grip intensive tests so this this glove may reduce discomfort and injuries from over grouping the photo that you see in the lower portion is a photo of the actual trial so we trialed it in body welds in our kentucky plant so the team member there is a is wearing the power pack which is a backpack and then he's actually using the glove on his right hand so it's it's not the best view of it but he's he's a he's using a hoist and he does other he does some other tightening operations too so uh the device looks like this essentially it's got a backpack with a power pack in it and then it has a cord and an arm strap that innervates that glove it's one power pack per glove and then it's got that remote control that the team member can adjust and it also is offered in a hip pack now so again this is an old photo bioservo has made updates to their product which i do not have pictures of but it is an ever evolving tool how does it work essentially it calculates the force that needs to be applied by the hand and assist the user by adding a portion of that force so essentially the the force that's required to perform the operation doesn't change but biomechanic way the person is applying less force because the glove is applying portion of it so that's basically on how to how to frame this uh in our kentucky plant we had eight team members they trialed for 10 days in fall of 2019 and they they wore a cut resistant glove or the iron hand glove as it is a required form of pee pee and weld shop you know due to the high laceration risk from those sharp stamped steel panels so the results i should also mention the iron hand glove does a digital assessment of the hand activity level which is another external tool through acgih and so here you have a bar chart that's noting with and without the glove in terms of that hand activity level for anyone who's not familiar with the hand activity level you see the scale there but it's basically looking at hand exertion risk of hand injury from motion and uh and burden so what we found we found statistically significant difference between using the iron hand and not using the iron the iron hand and in some cases using iron hand actually put the team member below the action level the action limit excuse me that the hands activate generates so numerically it looks really good in terms of reducing exertion now the feedback and of course team member feedback is extremely important and you know whatever we do if we decide to go any further with these devices or not um as i hope i explained with the shoulder devices so in the case of the iron hand the team members identified value in using iron hand but had pretty significant concerns with regards to comfort so one of them being the weight of the power pack um another big one being the wearing of two gloves so the iron hand couldn't be used in this particular shop because it's not cut resistant so having the two gloves made it difficult to manipulate small parts and then impacted tactile feedback so in some operations you do have to perform some flushness checks to see if there's any burr or anything in the panel and this was impacted by that particular device at the time there were some issues with the quality of the sensor so some fingers would stick some would not activate which i believe is a problem that has been teased out by bioservo at this point in time but again you have that negative feedback from from the team members so coming out of that and this is completely open to um discussion within and you know where we go from here but coming out of that very small localized trial this was our process decision flow chart should we ever implement this product in our operation so we would look for processes that are high risk for a hand msd as per internal tool t-bone that i talked about earlier processes that use some sort of jig or tool so essentially a high force power grip for more than 60 of the cycle time where it doesn't have a tactile check or or small part manipulation required with the gloved hand and no other control is feasible so same as shoulder exoskeletons and also we would need to make sure that the device doesn't pose a safety risk of any type so but again we're we're planning to do another trial to try to hone this process selection chart so the next step so bioservo has made as i alluded to earlier many positive changes to their product um we we tried to do a trial in 2020. of course you know kovid came along and it was really hard to do anything so um so we're planning to do a more in-depth trial for several processes in may again in our kentucky plant and we're actually going to be doing an additional emg study to independently confirm the reduction of effort so we're not going to just be looking at the hand activity level that is being provided by the bioservo we're going to have an add-on emg study so um for the for the back checking on time here so i got about 13 minutes so so for the back we certainly have some operations where we could we could implement some back devices so we have some operations that are lifting related or where they're engaged in static trunk collection so we are examining these pretty closely as closely as we have been able to do in 2020 with again the challenge of traveling having vendors on site covet et cetera we we did trial two devices early on and uh there were exoskeletons they were rigid frames for the back and there were there were very many trials uh very informal trials as in hey what do you think of this product and the feedback was was not that great so basically there was a long time to don and off so which is basically putting on taking off and they were uncomfortable because of a rigid devices they were difficult to move in some team members reported you know just pressure on the chest area which is particularly uncomfortable for our female team members and we didn't really go much further so but it's very exciting now because the images you see here are of very different devices that are emerging these days so what you see on the upper right hand portion is hero wear that is a back exoskeleton or exo suit that's really focusing on diversity and and women being able to fit women so which is really exciting and then the autobahn paxo and then seismic which the concept is you know clothing um or an undergarment uh so basically we're going towards the exosuit route as compared to the exoskeleton which is more rigid i really believe that her back that ease of movement will be really necessary for these items to be more easily accepted by people to wear so that is it on exoskeletons again that's really the the majority of our work but we're doing a lot of things in other areas so wearable sensors uh just jumping right into this so we we trialled something that we feel is really useful we're in the process of planning for a wider scale implementation so we trialled the strong arm fuse in 2019 and what is the strong arm fuse so in the photo here you see the team member wearing the harness and then the sensor itself which is that rectangular device on the torso area so it's a body worn sensor that uses machine learning to capture and analyze trunk movements to develop a safety score so if any of you are familiar with the lumbar motion monitor from ohio state dr bellamaris so it's not not to say this is exactly that but it's using a lot of the the concepts um from the lmm to a guide toward greater safety and lifting in addition to establishing a risk score this sensor is capable of giving haptic feedback or a buzz when the team member engages in high-risk exertions that are are set uh within certain parameters to encourage people to improve their lifting techniques the reason we're looking at it at toyota is certainly to reduce injuries for conveyance team members improve comfort reduce risk in general and also to reduce the amount of time to complete these risk assessments so we're looking at this for small law delivery operations our current tool that we utilize for these assessments takes about six hours to complete per process and it's probably done maybe two to three times each year per route we have about a thousand routes across all north america so if you if you do all that math it takes a lot of time to do these assessments as important as they are uh we can do these in some other more automated fashion and really focus people's thinking and skills on implementing controls to reduce risk that's really where we want to use that that people time so how this works um so you see a couple of images there so at the start of the shift you put on your harness and you go to the device there you go through the screen you pretty much looks like an ipad you scroll down you find your name you click on your name and a sensor lights up in blue and that is the one that you're going to be using for your shift throughout the shift the sensor automatically collects the data and at the end of the day you dock the center back into that docking station to charge and to obtain your personal safety score so essentially you get a score from zero to 100 and it's frames from uh athlete's perspective so what that means is that the higher it is the better you did so it's not framed in the traditional risk element so the higher the number the higher the risk so it's really frames from your point of view as in the worker what we saw in the indiana trial uh we we used this device for two months um so one month was just to establish baseline condition for team members and then after one month we actually turned on the haptics for 120 team members we saw 95 reduction in time to complete the risk assessments of course it's not a surprise so we're you know manually entering information as compared to using a sensor to automatically correct uh collect excuse me but the very exciting thing is that we saw 83 percent of team members improve their lifting techniques from the haptics alone so as measured by that device so again so that is is definitely very exciting so we didn't really have time to do much also it was only a month we didn't implement any controls onto the route which is something of course we would like to do additionally but only from the haptics we saw that improvement is really great and then overall we saw a 15 risk reduction by the end of the trial um really i call this a a game changer for us it's a very interesting new paradigm so the assessment tool is itself a control in offering that haptic feedback which is really something we've never seen before the the union of those two items assessment and control i'd introduced gamification into the process so we commonly heard the team members talk about hey what's your score hey i did better than you and kind of that friendly competition and and um this image here is really one of the things that's greatly won me over so um so it really supplements the process score with a personalized assessment so the fn 3a is the name of the conveyance route um so and traditionally our assessment tool you would have a single score um just hey here's the route uh it's yellow medium risk or you know low risk whatever it is here with the sensor we now have the ability to look at different people and how different people are moving and who is it possibly a higher risk of injury so the thought here is that we can we can work with those folks that are at higher risk really coach them to lift better of course also change the workstations and such but we can do a lot more one-on-one coaching whereas before you had a single process assessment and you were assuming that everybody was doing the same way but that's that's not actually the case once you start examining everybody so yeah there were some early concerns about monitoring and data usage was which basically went away uh throughout the trial so strongarm does have a safety pledge that you must not utilize fuse for punitive measures so it's really just about improving safety there were some concerns from the harness and of course female concerns about the location of this sensor on the chest and then we really need to understand more data to see if it actually does reduce injuries right it reduces risk but does that equate to reducing uh injuries as well and it was too short of time to really assess as it was only um it was only one month of haptics and again how long do we wear this uh and what happens when you remove the sensor how quickly does that good new lifting behavior extinguish itself right we know that from operand conditioning uh anybody familiar you know once you take away that feedback the behavior typically extinguishes itself over a period of time we are looking at some other centers as well so outside of strong arm fuse we're looking at a company called life booster where and then you see that lower right hand image that is a life booster so essentially you wear actual sensors on your body i believe you wear five and you get different posture angles we're looking uh we're trialing a watch that is called uh goex and that's looking at energy expenditure so it's looking at your oxygen consumption and a heart rate so we're trying to see if all these different variables can be utilized in our inner environments for for something if they're actually useful they can complement um the traditional ergonomics assessments and we're also researching industrial hygiene type of applicabilities so we do have some casting plants where there's uh where heat is a concern so we're looking at heat stress what we're starting to see is that a lot of these wearable sensor devices that primarily have the ergonomics purpose they're starting to build in some of these more industrial hygiene monitoring aspects so it'll be interesting to see how that goes in the future computer vision i'm going to start being pretty quick here because i wanted to cover this pretty lightly but computer vision we're looking at as well so computer vision essentially what that is allows computers to get an understanding from digital images um sorry i can't see my screen here but basically to automate tasks that human vision would traditionally perform so we've been working through this video sorry the video does not oh here it is i'll have to play it it does not play okay well essentially this would be somebody moving and you would see all these markers moving around with her and you have an image there of what that would look and that would be a moving image but basically computer vision you would not have to wear any sensors on the body so the team member would be completely free of sensors and just by using you know using my my phone i'd be able to capture some video and it would automatically calculate all the posture angles so that's another possibility of looking how we can automate a lot of these very lengthy risk assessments and make them more personalized for individual team members as compared to having that single process score so we're studying that we're also starting to look at some other areas outside of ergonomics so we started out with ergonomics but we're looking at some possible safety implementations so um some of our engineering groups are starting to use computer vision for quality inspections uh seeing if standardized work is being followed so we're trying to tap into into some of those aspects so for example computer vision could see as the is the team member wearing the appropriate ppe for the process is a team member trained and certified to use a particular tool etc so we're trying to expand outside of ergonomics as well for computer vision we're looking at virtual reality as well so virtual reality so the background for this so are acute injuries so these are things like lacerations slip strips and falls et cetera so again this is outside the realm of ergonomics our top root causes for acute incidents are the ones listed here incorrect method followed lack of awareness and improper training so we're trying to utilize virtual reality to enhance our safety training so basically the theory is that you you retain based on your level of involvement so what we've done traditionally is focus on passive training techniques so basically an e-learning or a video or an instructor-led training we haven't really put people right then in there to experience things so we're trying to do that through virtual reality we're focusing on initially at least anything can benefit from virtual reality but uh ergonomics here is not really our top priority for the time being so we're looking at high-risk tests that are difficult to simulate or high retention is crucial so we're looking at those more acute um very high risk incidents of possibly lockout cranes with heavy overhead loads electrical safe work power gesture vehicles etc it's very exciting so the world of vr is becoming more and more uh interesting so there are emerging accessories that um again are going to really provide a whole other aspect to vr so here you have see a couple of examples both of these are going to add additional dimensions other than the visual to vr so that vr glove that i have pictured there and acts tactile feedback the tesla suit on the right hand side and there's no relation to tesla the auto company it just happens to be the same name so it's going to use electrical muscle stimulation to make you feel pressure vibration weight heat cold really exciting the things that we could do with vr training with help helping training with tension there is a different group that is examining digital assembly and virtual reality very closely um so we are not doing that at this moment in time but they're looking overall how to use vr to build a vehicle so the images that you see are of uh the traditional program that we use to see how we can build the car safely and ergonomically and we put a mannequin in place there to see how different things can be done but we're not experiencing them directly so we're hoping to enhance that through vr as well and we're trying to understand how to use vr in a manufacturing environment so um you know on the lower half of the page you see somebody gaming at home on the upper side you see a toyota plant so again it's it's very different the concerns that pop up uh in terms of how to do this so a significant concern again if you're aware is cyber sickness so essentially that's motion sickness that's induced by virtual reality some of us are more sensitive than others and so we're trying to understand how do we compile these vr training enhancements without actually giving people motion sickness as you could imagine they had to go back and work on the line after so we're definitely very concerned about their their safety and the last bit that i will touch on because i'm running out of time is artificial intelligence uh artificial intelligence i'll quickly define it it's the theory and development of computer systems able to perform tests and normally require human intelligence so i think this is a really good way to frame what ai is so this is from ai summit i'm not an ai expert but essentially how they explain it is that you know first it was you know essentially paper and ink humans were not optional and we are really primarily in the age of software so where tasks can be automated so repetitive and prescribed tasks can be automated without humans and we program software the ai realm is really going to be where human reasoning begins to be automated and we coexist with machines and we do not program it we actually train ai to to help us some daily examples of ai you may be familiar are listed there siri amazon etc uh retail industry is a huge user of ai so i think i read a figure like 30 something percent of amazon.com
purchases are actually generated through suggestion engines based on previous purchasing behavior so retail industry is a huge early adopter of ai and this is the growth of ai here so again it's currently evolving retail is the one primarily using it in the next 10 years some experts say that again is not for me this is from experts uh the conference that i attended called ai summit if anyone's interested it's going to be growing and targeted applications medical and legal and then you know 10 years beyond it's going to be more human mimicry so machines are going to behave and sound more like humans and then beyond that they claim human consciousness which might be a little scary thought hopefully it's not the age of the machines but that is the progression so basically how do we tap into this how can we use artificial intelligence to assist with with reducing injuries our plants i don't have any images to show you here a lot of it has to do with incidents which which i can't show but essentially we do a lot of prioritization and risk prediction based on things that have happened in the past not just risk assessments but also where have injuries occurred so we do that manually now people spend a lot of time on this so we're trying to understand how artificial intelligence can help us make those connections a lot faster by joining different information which was maybe previously siloed and then this is really my last slide here so i am a little bit over but i've showed you a lot of different things that are maybe don't all go together right now and this is a really big question that we have so technology ecosystem how do we effectively create this so ecosystem being how do these things all live together so we don't want to put 50 devices on people so how do we build this as an ecosystem where all these different assets have value and can coexist naturally and comfortably for people and really get the utmost benefit so i know that was a whirlwind uh talked really fast towards the end so i really appreciate your interest and attention in this and feel free to connect with us if i don't get to answer your question within the seminar i'm happy to connect further through those methods noted there that's all i got thanks thank you marisol i had unmuted brand but i'm not sure exactly where let's see maybe i lost him he's he looks like he's mute he's muted let's see if we can do this can you hear me there we go okay thank you jessica thank you marisol for a great presentation and with respect to time we're going to segue into some of the questions that were posed so i apologize if you're if your name shows up in the chat box i will use your name if not um or i can't differentiate i'm going to use your username that was provided but marisol we had a question from chris reed and he just wanted to know if you could explain why toyota would make some areas or types of work mandatory in other areas voluntary and were there any qualitative or quantitative reasons for that reasoning so why required in some areas i i think it just has to do with the with the level of risk um so everybody's familiar with other types of ppe or most people are so in terms of you know safety glasses safety shoes hard hats you know respirators etc anti-vibration gloves etc where where we see that risk being there so in this case the overhead work or some other possible risk in the future outside of that if we see that risk that can be effectively reduced with a new device and it doesn't pose other you know significant concerns of you know comfort uh other safety issues etc again it's really in the world of safety it's it's for us it's really for me uh speaking for myself it's almost a moral responsibility to really provide that um and and have people use it so um so again we're trying to make it so that everything is as comfortable as possible but i assure you that most people also do not like to wear safety shoes safety glasses hard hats etc these days most people don't like to wear face coverings either for covid but again it's it's things that we do that will provide us with greater safety so and that's why we require it just the same as all those other devices with regards to the voluntary issue we we have found some team members that truly just want to use it even though we haven't seen such high risk they're really interested in using it and we're we're making that an option for them um so you know a lot of people look across the the line and and see somebody else using the device and where it may not be required in their process but if they see value and they feel value we want to be able to provide that as well even though it's not required so that would be the differentiation hopefully that makes sense okay thanks uh we had a question from jake and he was just wondering what was the decision that led you to try the product with the harness and if there were any pros or cons with respect to the harness versus different products [Music] yeah that's that's a great question so there are other sensors out there certainly um so this particular one from strongarm fuse i think they've done a a great job in being able to provide the information in a very simple manner to the team members who may not have a strong safety or ergonomics background their center collects a lot of data but then compiles that through their proprietary algorithm that i don't have it either but and it gives you a single score that anybody you know who's ever taken a test in school or ever has gotten a grade can can relate to zero to one hundred it's telling me how i'm doing so um just that aspect of it really resonated just being it having it be really simple for folks to understand not just safety professionals so that made it really attractive uh the harness had was a concern as i understand it strong arm is actually developing different ways of um of tethering that sensor to the body including a backpack so when we when we do a trial hopefully soon in 2021 at our indiana plant uh we will be giving team members the the option of utilizing that device however they want to use it and i think they're working on other not to speak for strong arm i think they're looking at other possibilities of um having that sensor on the body such as a you know even like a t-shirt with a pocket um things of that sort sounds great so another question from vx609d the question is how was or is the employee's response to the required ppe yeah i mean i'd i'd definitely be lying if i said everybody's like incredibly happy about it right so we had we had similar responses to again face coverings other pp i think the plants have done an incredible job of communication um you know if if you just come out and say hey you know next week we're going to make this required um and you know that's that that sort of thing does not work so it's a very long process where you start sharing the information the why is so important the why of what you're doing something is is so critical and how it's put together that communication piece so really showing right we're concerned about the safety we really want to improve these operations we realize we can't improve it in other all these other ways so we want to do this and really that communication is really uh important so i think a lot of folks that had concerns early on have come around um there are some that you know maybe don't accept it and maybe just put it on because they have to and are maybe not so happy um but again that happens with all people so we're really concerned about safety while we still work on making things uh as comfortable as possible for everybody sure thing so next question from s-a-t-y-a-j-i-t they were wondering if it took any time to get used to the exoskeleton and how did you train employees using the exoskeleton and if so how was the training performed yeah so uh another really good question so again our plans i'm i'm really sort of the voice in this level for a lot of our amazing plant safety professionals that really do this hands-on but it's about the introduction again so there's a ramp up there's a there's a whole schedule that we have in place with introducing these devices so let's say there's a new model change and there's a you know set of new processes the assessment is done and the assessment is flagged and we go through the process to designate it as a required exoskeleton process there's the whole you know communication speaking to the team members introducing the device fitting them appropriately and then there's a ramp up process so when they start wearing it um you know they only wear it for a very short time the first week then the next week they ramp up a little bit more and they start wearing it a little bit more you know hours per day so again it's a it's a very long introduction process that whole time they have to be refitting addressing the team members concerns hey you know i need to kind of adjust this buckle um i need to maybe lower the resistance uh et cetera so it's a it's a long process and that's why it's you know if you're just gonna take these devices and kinda hey you're all wearing this because you know we saw toyota had really good impact that really doesn't work it's a whole program that goes along with it like anything else thank you so from jake he had a question and asked about the future application for vr training to prevent slip trips and falls what do you think [Music] yeah whoever that is if you have some ideas i'd love to see that so uh slips trips and falls are a big issue that we have across toyota like in anywhere else i think lots of manufacturing so i'd love to hear some ideas so we are developing um a safety critical eye training in vr that will help team members spot uh hazards more easily uh slips trips and falls for us i mean they they occur for the main reason is uh 5s again it's basically having a very clean and orderly environment so some you know parts that shouldn't be on the floor somehow get on the floor um you know something gets on the floor that shouldn't be some sort of spillage etc so being able to spot all those things more easily that'll that'll definitely foster less slip strips and falls sounds great so the last question uh that we probably have time for is from jeff oil and he just was wondering how do you handle the sanitation and the cleaning of the exoskeletons yeah um so this past year we we didn't actually see uh an impact from uh from covid for the exoskeletons that were already in place um so basically each team member gets his or her own device that is not shared so they would clean it just as they would normally clean it with an antibacterial wipe etc some of our sites do have some devices that clean through ozone such as our indiana plant um and uh they they do effective sanitization through that we haven't seen an impact there but we have seen an impact we're trying to get um new things in and such so with a lot of our trials like vr that's probably the as you can imagine putting a headset on is a very significant risk of covid so we've developed a pretty rigorous protocol uh requiring ppe to be worn so for example those i forgot what they're called right now but basically there are little eye masks that protect you from the device and of course alcohol wipes etc using even hair nets and gloves while using the devices and in addition we've invested in some devices uh they're called a clean box it's uh essentially a way of sanitizing the headsets through uv light uh so we're doing that as well in addition to all the other protocols so it's been challenging to do new things uh due to covid for sure we're still trying to go on again the injuries injuries are not going to stop just because there's code so we really need to look at these promising options to help us in that room we actually do have two more questions uh quick questions um so ryan porto asked are you having more success with implementation with the sensors or computer vision in general so yeah so i i was really hoping ryan i was really hoping that computer vision was going to be the key to everything so it sounds wonderful you don't need to put anything on anybody you know you just film and go and there's your information right it is proving more challenging than sensors because we're a key item that we've had is really developing the infrastructure to be able to process that info super data heavy we've had to set up servers to actually be able to process that information which is very complicated and such to store and hold that information interpreted accordingly so um we're really going to see how this comes out and if it's something that we want to pursue further in that area but right now the sensors the sensors are a lot easier um now of course you know there's there's some cons to them right people have to wear things right being the number one right um so again it'll be interesting to see and i think time will tell how we can integrate maybe the two items or really just choose one over another until maybe computer vision works a little bit better or we can make it work a little bit better thank you so the last question is from jgb513 was there an opportunity to determine roi related to lost cost and production improvements when adopting the various technologies yeah um so i'll i'll note the uh the strong arm fuse sensor as a key example so uh we the reason we are we're implementing it at our indiana plant um as soon as it's okay to do so uh based on cova uh this in 2021 so we actually again in terms of applying these things you have to really look at the dollar amount for everything so we were able to look at um you know the injuries that we have had in those areas and the possible improvements based on the risk reduction that we saw on the trial we were able to put a number to that so again it's about people it's about safety but again
2022-04-19