Case Study: Virtual Reality Training at FL Techncs

Author: Ramunas Paškevičius, Head of IT and Innovations, FL Technics

Subscribe

Ramunas Paškevičius, Head of IT and Innovations at FL Technics explains the Implementation of Next Generation Mechanics Training using Virtual Reality (VR)

In this article, I want to share our experience at FL Technics of using Virtual Reality (VR), something quite new to MROs. FL Technics is still at a very early stage in understanding how this technology can help us but, with what we know already we strongly believe that it will work. One thing that led us to this conclusion is the expected future requirement for technical staff as illustrated in figure 1.

Figure 1

The map predicts that, in the next ten or twenty years, there will be a dramatic increase in demand for MRO mechanics. What that means for an MRO organization like FL Technics is that we’ll need to prepare increasing numbers of mechanics with the need to attract people from different industries who we’ll need to retrain. Readers will understand that training or even retraining an engineer or skilled mechanic takes a lot of time. In FL Technics’ case, in order to get what we call a level 1 mechanic authorized to work in our hangars takes up to six months. However, while customers might not wish to have a trainee working on their aircraft for those six months, we will still need to be paying their salary. So our objective was to find a way to reduce the training period to weeks rather than months but with no loss of quality and we believe that virtual reality will help with that.

FL Technics is expanding very fast from its origins in Lithuania; not just in numbers but also with its growing global presence (figure 2).

Figure 2

Established in 2013 with just one hangar at Vilnius International Airport in Lithuania, in 2014, FL Technics opened a second hangar in Kaunas, also Lithuania. By 2016, realizing the potential of the Asian market, we established a facility in Indonesia with the view to transferring our knowledge, experience, quality and standards to the Asian market. One of the biggest obstacles was how to work with the local people maintaining our very high standards. We know it’s a different culture but we needed to achieve the same results as in Europe. The most recent project, which has been very successful, is to open a hangar in China at Taiping International Airport. We see these steps as just the beginning for the expansion of FL Technics but it further reinforces that we’ll need to standardize training for mechanics; maintaining the same quality and reducing the training period.

In all this, our focus is not only on maintaining aircraft but to aim for the highest quality as can be seen from our customer satisfaction results in figure 3.

Figure 3

We monitor our customer satisfaction index each year and it’s increasing. The standard in Europe is for 60% to 70% customer satisfaction but we have now lifted our rating to around 80%. Customers choose to work with us for the quality of communication and innovative solutions, bearing in mind that FL Technics is an independent MRO so with no base load of aircraft from a parent airline. We need to compete for each customer and to retain those customers; there is only one way to do that, deliver the highest quality.

Our latest achievement in the last couple of years has been that tier 1 customers like the Lufthansa Group bring aircraft to FL Technics for maintenance instead of using their own facilities. That tells us that we’re doing the right things in the right way. To maintain that in the future we’ll need to find ways and opportunities for utilizing new technologies. It’s no longer enough to work in old-school ways including with training. Plus we also have a Part-147 organization as our training organization where instructors still go into the classrooms where there are still old paper training posters. That has to change and we need to move forward; but the first consideration when looking at the virtual environment is its availability.

OPPORTUNITIES FOR USING VR IN TRAINING

Whether we have 20 or 50 instructors, they can support only 20 classrooms at a particular time which is limiting on the space available for training and limiting on the number of students who can be trained. The first thing that comes to mind is how to show students what can happen in a real environment in risky places. For instance, working on the emergency doors or something like that, there is a risk that if things are done in the wrong sequence, an explosion of the escape slide can be triggered within the hangar (figure 4).

Figure 4

At best, that will be a financial loss, at worst it will injure the mechanic and other mechanics. We did not have such cases in our hangars so, while all of our staff know theoretically that there is a risk, they have not seen how that explosion and deployment of the escape slide would look if it happened in the real world. Virtual reality is able to show mechanics that event and for them even to feel it.

The procedure in the demo video at the start of this article was selected for a couple of reasons one of them being the big risk: if the thrust reverser is opened in the wrong configuration, you can damage the reverser, which will cost something like €100,000. So, our mechanics know that there is a risk but they have never seen how it would look in real life. A virtual environment allows us, during the training, to show the consequences of that mistake.

Next, if mechanics are able to be available for training 24/7, it means that there is no need for an instructor sitting next to the trainee and explaining the steps and procedures. The trainee can go inside the virtual environment and undertake training processes. Again, the example in the video is a real case in FL Technics’ hangars. The procedure shown is performed ten or 15 times in a hangar, so if there is a need to get some real experience, then the trainee will need to wait until a customer orders that task, then go to his or her supervisor to request to see how this procedure looks. However, the trainee won’t be allowed to do a physical execution of the procedure. It takes time. In a virtual environment, the trainee can put on the headset and practice the procedure or task hundreds of times.

During the time when people are training, a lot of data can be collected. Current technologies in a virtual environment allow the tracking of many things. The individual trainee’s movement, their reactions, their decisions, exactly where they are looking… This information can then be used to, for example, increase the quality of training material. So, if you, say, get data or can see that a trainee is looking for a particular tool but doesn’t know what it looks like, then the training provider might make a decision to add material to that specific training to show what the required tool looks like or where to look for it. It gives insights which would not be possible in a traditional learning environment because, as an instructor or training designer, you’ll never get feedback from the mechanics you trained and are now working in the hangars.

The main goal that we were seeking at FL Technics was to reduce the time of on-the-job training or even to eliminate it. That’s an ambitious goal because there are authorities who are not keen to look at these technologies to replace on-the-job training in a virtual reality environment so the person who’ll get the experience will not actually touch a real aircraft.

In that context, the latest thing with this technology is automated testing. The system or artificial intelligence (AI) can look at a trainee’s behavior in a particular situation because virtual reality could simulate some things. Now we’re working on one new topic for the general visual inspection where dents will appear randomly on the aircraft, and the trainee, if they’re new and have never before done that, needs to find them and identify what they look like. The system will be able to analyze how the mechanic looks for the dents, where their attention point is, how they perform, what tools they use and so on, and will provide a result. We believe that this technology will work because there is nothing new in how it operates (figure 5).

Figure 5

Most readers will be familiar with the pyramid in figure 5, which tells us that, using just lectures or training materials, retention rate for what has been learned will only be between 10% and 20% maximum. But using VR fits into the level of ‘Practice by doing’ which has retention rates of 70% to 80%. This is based on standard material  but a couple of MROs are doing experiments, testing and measuring how this will all work in the real world and say they have already achieved retention rates between 70% and 75%.

What does it mean for the training procedures? It’s possible to reduce the time spent in the classroom to a third or less. It means that you can get very small theoretical subjects covered in a classroom the spend hours in a virtual reality environment to get the same results

TRENDS IN VIRTUAL REALITY

Returning to the technologies and the main items that are available or coming to market, we did some research at FL Technics which revealed some ideas that I would like to share with readers (figure 6) about how the technology will evolve over probably the next two years.

Figure 6

First of all, from our experience, if you want to use virtual reality based training, the main requirement is wireless technology and devices. We tried a headset with a wired connection and found that it limited the user’s movement. After that, we switched to wireless headsets which offered many more opportunities. So, in our case, we have tried in one case to use a simulation in a place that was about 30-40m² and where the user can move freely. So, if the VR user is standing next to an aircraft, it’s possible to go around the engine or around different parts without the need for, what is called in virtual reality ‘teleporting’; the user can just find and feel the scale and characteristics of a real object. I think that all simulations, all the training in the future will happen only on wireless devices.

The next thing is what we discovered from our experience and from the people who are using VR that if you have used VR at home for entertainment simulation, it usually includes a couple of things. First of all, there will be sensors; there will also be a quite large and powerful computer and transmitting devices which, together, make up the full infrastructure. And it’s not very mobile: if you want to move the classroom to another place, it takes time to do that. In light of that, we strongly believe that the device used should be completely mobile, i.e. no need for a workstation, no need for sensors, everything is in one place, one device.

The other point is that to achieve a good feeling in a virtual environment there need to be at least six degrees of freedom so that users can move forward, backwards, left, right, can tilt and can make other movements which will all be reflected in the VR image. This will ensure that the experience is as realistic as possible. Feedback from FL Technics’ mechanics was that, if it is like a game with fixed pixels and limited movement, it will not deliver a realistic environment. There are now some evolving technologies using not mobile phones but a platform based on the mobile phone platform and, I think, in the future that will be a common thing for one overriding reason, the price. For example, good virtual environment equipment will cost around €3,000 to €5,000 a set. So if you need ten or a hundred sets, the cost will be high which, in turn, will become an obstacle to adoption. Oculus Quest (figure 6 above) costs just €300 while offering almost the same quality and the likelihood it will improve with time. So, it is likely that all virtual environments will be mobile and very cost effective.

Another issue is that the headsets are not comfortable for use over extended times during a day and can be painful at touch points. Most vendors are working hard to improve comfort so that, in future, helmets will be so ergonomic that users will not feel them and will not, even after up to six hours in that virtual environment, suffer any backache or headache. The last but very important point for collecting data and conducting examinations, is tracking. I have been surprised by a couple of new technologies that are in the pipeline or already existing for headsets that they can track not only where the users moves but also how fast and with what force, plus track a user’s eyeballs to see where they are looking. So, if the material is improving, or giving a mechanic a situation where they need to do troubleshooting, the system can track where he or she pays attention, i.e. looking at the right point or how they find an issue in a particular situation. Collecting the data and doing the right analysis will be essential for training, especially to convince the authorities that the entire life cycle of the training is correct.

CHALLENGES IN VIRTUAL TRAINING

Looking at the challenges that FL Technics faced; the first one concerned quality. When we asked the quality team for a path by which we could deal with the authorities because we wanted to replace the on job training for the mechanics, unfortunately, they did not really understand what was wanted. This is a big issue but there are a couple of other MROs working to get approvals not only from their local authorities but from a wider group. For instance, we are speaking with EASA to ask them what to do to get approval, what do we need to provide or what evidence would we need to produce in order for them to have confidence in VR-based training. This was the first challenge on our journey to virtual reality: the second was FL Technics’ management. When we first put the proposal to senior management, having completed a pilot project and having a training scheme mapped out, their first question was ‘how much will this cost?’ And when we explained that the pilot cost around €20,000-30,000, for one task when just one aircraft type can require 800 tasks, they did their calculations and replied in the negative. They did not wish to build another hangar in virtual reality, it would be too expensive. So, there are a couple of approaches on which we have been working to try to provide the right numbers.

The first one is probably to find ways to co-operate with OEMs which can seem impossible but we need to talk as well as creating alliances with other MROs on developing content because the 3D and interactive content is the most expensive part of this. In FL Technics’ case, we built a very precise copy of the hangar so that users feel that the virtual environment is the same as the real world environment that they’re used to. Content sharing might be a way to achieve lower costs plus there are a couple of things we discovered. It is not necessary to have all tasks in the VR solution but it is important to find particular tasks with which there is a high degree of risk, such as the already mentioned deployment of escape slides, or thrust reverser or other tasks which can damage the aircraft. In such cases, money for VR is not a problem because, if an aircraft is damaged during maintenance, it will be necessary to delay work until it is fixed and customers will not be happy to be told that delivery will be two days late because we made a mistake. In a case like that, it’s not just about the money but about our reputation.

The second thing is how often a particular task will appear in the hangar. For instance, if it will be a year before the mechanic will get the chance to work with and gain experience of a particular task, the MRO will need to continue to pay him or her until they get the chance which will be a significant cost for the business. So, these will be key points in discussions with management on how to overcome the concerns about cost.

The main challenge that we faced, though, was to engage people to use this technology. In FL Technics’ hangars, there are two types of people: the older and experienced ones who have been working for twenty years or more in MRO, and the younger ones who have only just been certified. The young ones are keen to use the technology and actually enjoy it. The older ones are more resistant to VR, wanted to continue to gain their experience on actual aircraft: they did not care about the cost but simply wanted to do their job, maintain aircraft. So, without engagement, nothing will be achieved because when you build a realistic VR model, you need feedback from users.

In the case of FL Technics for this pilot project, there were three certifying staff people, one person from Quality Control and around ten mechanics to collect the feedback from them. The Quality people pointed out those virtual elements that didn’t look very realistic; the mechanics said that some actions such as how they held their fingers to open latches were wrong or the device’s feedback was not in the right place on the screen. As a result of those feedbacks, we addressed all of the concerns raised. Without involving the users and others around the business in any creation of training material, it will be impossible: even IT experts and training instructors will not be able to create material without these feedbacks.

Finally, about the technology: there are a number of businesses involved in the virtual reality and gaming market. We had to choose between the 3D engines and our idea was to go as realistic as possible. In that, we went with Unreal Engine which, at a certain point, was the best at recreating 3D images. As a result, we have an 8000m² hangar in virtual reality even with marks of wear and tear on the walls and on the floor… everything is copied as in the real hangar. It is important to have something this realistic so that people feel it’s nothing special, everything is what they would commonly see in real life.

Our experience also suggests that it’s best to carefully choose the partners who are producing your content. We had two opportunities to go with cheaper ones, students, or to go with professionals who have been in VR for a couple of years or more. We sent them the same task to create and the students did it much cheaper and faster but it was really like Minecraft, not as realistic as we wanted. Against that, the professionals delivered something that was 95 percent realistic.

LESSONS LEARNED

If readers are planning to do something in their shops, hangars or even training for airline personnel, the first thing is not to start programming things but contact your local regulatory authorities and quality regulators at the very beginning. Tell them that you plan to implement VR-based training and ask them what they need to know before even the start. If you include them in the team, it will run much faster. Also, when we involved quality managers only at the end of a process, we found ourselves having to redo 50 percent of the work after the quality people pointed out what was wrong, what went against our procedures and that extra work cost a lot.

Another lesson is to search for repeatable processes which make sense of moving to a virtual environment plus you need the right people. It’s important to attract to the project people who are willing to make organizational changes and who are innovative and passionate – they want to change the world. They’ll be engaged and prepared to work as much as is needed because they’ll enjoy participating in the project instead of being pushed into the work. Finally, it is important to select the right partner, a good partner is essential, not only as a supplier of a virtual reality solution but who is committed to the success of your project. The partnership should be one of equals: when we had a supplier customer relationship, the supplier would finish the job when they had completed what was in the specification notwithstanding any problems that had emerged during the process. But, when dealing with virtual reality, nobody knows how to correctly specify the logic of the environment. There is no prior experience, especially in the aviation sector, how to do things correctly when specifying VR.

NEXT STEPS FOR VR IN FL TECHNICS

So, where do we plan to go from here? First of all, in Lithuania, FL Technics is building an engine repair shop and we think that an engine shop would be able to deliver a positive ROI from virtual reality training because, if you want to train a mechanic on engine repairs, you need to buy engines. And, with another shop planned in Asia, there will be the need to buy another engine because there needs to be in a training facility next to the shop. We believe that VR will help us since we perform engine maintenance on a limited range of engines and engines are not as unique as aircraft so it’s easier to create standardized training for mechanics.

Also, thinking back to the start of this article where we looked at future demand for people, and that, in future years, we’ll be attracting people from different industries. We’ll need to provide them with very basic training about simple things in the sector such as that some composite panels must be washed by hand, not with a pressure washer which will distort them. There are lots of examples like that and so we think there will need to be a three step program:

1. A newcomer attends a theoretical course;
2. Then they go to a practical level in order to become familiar with basic tools;
3. And finally they go to VR to go around an aircraft and perform very simple tasks.

That, we think, will be the beginning for a real life application for virtual reality. It will not cover heavy maintenance yet but will be a beginning.

Our goal is to replace some OJT (On the Job Training) with virtual reality so we’re working with the authorities to determine the acceptable path to what we want to achieve and I sense that even EASA is being supportive of this program, telling us what we need to change and what proofs will be needed. With this in mind, we found a local university with a scientist whose background is training so that we can undertake studies with a scientific background on real and virtual environments testing to get the results that will satisfy the authorities that the quality of training will be the same.

All of these things will enable us to move forward and be successful in this new style of training.

Contributor’s Details

Ramunas Paskevicius

Prior to joining FL Technics, Ramunas Paskevicius worked with private and public sector IT projects. Since 2011, he has led FL Technics team in developing, integrating and supporting a wide range of IT solutions for MRO, sales and administration. With a Bachelor’s Degree in Business Administration and Management as well as a Master’s Degree in International Trade, Ramunas understands what businesses expect from IT and how to ensure smooth integration of IT solutions and engagement of users.

FL Technics

FL Technics is a global provider of aircraft MRO services. The company is a family member of Avia Solutions Group, the largest aerospace business group from Central and Eastern Europe with 67 offices and production stations. With over 20 years’ experience in the field, FL Technics specializes in base and line maintenance; parts and materials trading; engine, APU and LG management; full aircraft engineering and design, and technical training services.