Aerospace is a growing, changing industry. More information is being collected, creating more data that needs to be analyzed and extrapolated to continue to innovate and advance a company in this industry. Companies want to create aircrafts that are more efficient, more customizable, better designed, faster, cheaper, safer, and more environmentally friendly. To do so, all of that information needs to be processed and understood.
Hundreds of pages of technical text can be difficult to read and organize, which could lead to an important requirement being missed. As the industry grows and changes, new solutions to the problem of text data are being created.
Changes, data and regulations
Sensors onboard aircraft are constantly collecting data. Information on the systems, equipment, and current and future weather conditions are sent from aircraft to pilots and ground operations. In the near future, planes could be able to minimize turbulence and maximize fuel efficiency by altering their routes using real-time data. Autonomous flight is getting closer to reality.
Using analytics to improve maintenance efficiency is increasing in popularity. Predictive maintenance uses analytics to determine when to replace certain parts of an aircraft. This helps maximize the lifespan of these parts, ensuring safety and minimizing equipment checks.
The space industry is getting bigger and more varied. Commercial flights and cubesats are providing a completely new business model and service design for the industry. New technologies are being introduced, making the engineering projects more complex.
Customer demand and the goal of achieving sustainable growth are factors for growth and change in every industry. In aerospace, there is a new focus on reducing emissions, innovation in cabin design, and a reduction in travel time. Production line efficiency is a way for companies in aerospace to achieve sustainable growth. These factors are reshaping the industry, changing how companies operate.
This increase in complexity is providing more and more data that needs to be processed, analyzed, and implemented. With a lot of safety critical aspects to take into account, text data for safety requirements, engineering requirements, and new regulations are expected to increase. With all of these changes, there will be even more text data to process and organize to ensure all regulations and requirements are being met.
Issues with processing vast amounts of text data
During procurement, time can be of the essence, and reading through hundreds of pages of technical text takes a lot of time. A large document may need to be read and understood very quickly so that a new project can get underway. Unfortunately, this can lead to some parts being misunderstood or missed entirely. Leaving room for error is a huge risk to the company, and can put them in breach of the contract they’ve signed, not to mention the potential for grievous errors in the finished project.
Engineering projects in the aerospace industry are based on a set of written specifications. These text documents are typically very large and can be difficult to navigate to the desired or applicable sections. They are also very complicated when the reader is trying to extract only the sections relevant to certain groups.
For each new space mission, a very large set of written requirements is passed down the supplier chain, all of which needs to be met and verified. Missing a relevant set of requirements can complicate projects and impact the project’s bottom line significantly.
With all of the advancements in technology, the aerospace industry is changing at an alarming rate. Devices are improving, becoming more complex, and increasing in popularity. All of these changes are creating more data in the form of requirements and regulations that need to be processed and analyzed before a project can be completed successfully. Trusting this job to a group of people who are assigned the task of reading through hundreds or thousands of pages of technical text can leave room for human error.
What would happen if a company missed a new regulation put forth by a governing body? Or if they didn’t catch an engineering requirement that directly contradicted a new safety requirement?
People can misunderstand requirements or fail to catch contradictory requirements when given the task of analyzing vast amounts of text data. Human error is not always something that we can afford. Fortunately, however, there are tools that can help alleviate these risks.
Solutions to help with processing text data
Where Ansys Esterel helps with critical system code, there are a set of tools that can also relief complexity on the text data and requirement side.
Skywise, by Airbus, aims to connect all aspects of the aviation sector, and eventually all of aerospace, by allowing companies to improve efficiency through the use of one digital platform to integrate all data sources. Compiling shareable information on maintenance, safety, and other workflows allow for greater transparency and increased communication and efficiency.
Selko Analytics uses AI to help locate specifications specific to certain groups, using “Intelligent Search” for a fast analysis of the text data, or user-trained machine learning that requires only 100 samples.
Using Artificial Intelligence to categorize and analyze vast amounts of data ensures that a person can spot risks, important requirements and regulations. All results can be verified by a person before export, and any changes made will be learned to produce better results next time.
Surrounded by today’s complex technology, it is important to stay on top of the game. Solutions such as these helps reduce risks and help manage all that text information. They help companies optimize their efficiency in an industry that is very quickly changing.
Selko team present at the SEAnnovation area hosted by Starburst at this year’s Euronaval in Paris. The first day has flown by with Selko’s CEO Tuomas Ritola pitching at the startup stage, and great conversations with European engineering giants.
We are still here tomorrow, so come visit us at stand 14 C10/E17!
Maritime engineering often involves large development project, and safety critical aspects are often at the center of consideration when developing something new. This results in large amount of technical requirement text that needs to be processed and understood.
Looking at the expected changes in the industry in the next few years, it is safe to say there will be an even larger amount of regulation to deal with. Engineers already spend a lot of time processing this text, and this is likely to increase in the future.
Regulations, standards, customer requirements…
Depending on the project and the field, developing something new, or simply replying to a customer request in a new location, can include a lot of extra work when regulations and standards are involved.
Regulations change from region to region, and the International Maritime Organization (IMO) works with governmental regulatory bodies to oversee collaboration in the field of shipping and international trade. IEC and ISO also play a role with several field related standards, such as ISO/TC8, ISO/TC10, IEC TC 18 (and many more depending on the field), which are considered together with regulatory bodies to avoid duplication. For ships, an extra layer of complexity can also come from being able to conform to a certain class.
Engineers have been battling with regulations and technical standards for years, but an important source of textual requirements comes also from customers and stakeholders. When selling a complicated system, or when purchasing them, there is a lot of text data that needs to be processes by the experts. This can be a matter of working through received tenders, or separating pages upon pages of customer requirements to the right teams. Simply put, the more complicated the system, the more data is often involved.
Complex in nature, large in amount
With large amount of data comes also large amounts of difficulties. Communication between different teams in a projects can be hard if the information is vague or ambiguous. All requirements need to be verified, but with thousand of pages, it is easy to miss the important detail. Information coming from many different stakeholders can cause contradictions in the analysis and it can be hard to spot these issues before it is too late. If the text corresponds to a longer project, any issues not solved in the beginning of the project, can have exponential costs down the line.
New technology showing the way
The amount of data is only going to increase. There is an incomprehensible amount of data generated at any given time; by 2020 this is expected to increase by 4,300%. This is not all without cause. Looking only at the maritime industry, several radical changes are taking place; autonomous ships are being introduced, sensors are being applied more than ever, robotics are helping with complex tasks and communication systems are expanding.
With new technology come new regulatory issues. Safety critical systems are getting more common and this will cause an increase in requirement text and need for verification.
Taking advantage of this huge amount of data requires a different way of thinking: How data is built and used? How much of it we can handle? How fast we can process and analyse it? And where and how are decisions made?
How to deal with all the information?
Luckily there also exist technology to help deal with this. For managing technical requirements solutions such as IBM Door, Polarion and Jira can offer a platform for controlling that information.
On the procurement side, large integrated tools such as SAS Ariba can help control a large set of suppliers and improve supply chain management.
Artificial intelligence can also be utilised, and tools such as Selko Analytics can automate some of the tasks in processing large quantities of text.
And finally, to keep it all together, protocols such as the Shipdex Protocol can help keep everyone in sync.
Maritime engineering is a fascinating field, that is facing some exciting changes in the coming years. People and organisations will interact differently and processes can be improved, leading to better and more rewarding work. Being able to harness this data and get the most out of it is going to be more important than ever, but also present a great amount of opportunities.
We are talking today with Paul Beecher from the World Economic Forum.
We will be talking about trends in production, how new technologies are shaping the world, who the key stakeholders are, and how policy can be used to aid less developed countries grow.
Paul has been researching policy and manufacturing, first at the Institute for Manufacturing at the University of Cambridge, and now at the World Economic Forum in Geneva. Paul was recently a project lead for the “Shaping the Future of Production” System Initiative.
Paul, you’ve been working with the Institute for Manufacturing at Cambridge and the World Economic Forum so could you tell us a little bit about your background and how you got into this?
Well, I started off as a research engineer back in the early 2000’s, mostly in nanoscience. Researching novel nanomaterials for electronics, be it metal nanocrystals, carbon nanotubes, organic semiconductors, inorganic semiconducting nanomaterials, graphene, etc. And I suppose a time came when I had an interest in having a broader purview.
I was interested in how strategic decisions get made, I was interested in this notion that in, advanced western economies especially, the knowledge economy is very important, and that led to my current role here in Geneva with the World Economic Forum where there is an initiative called Shaping the Future of Production and I’ve been working on that since late 2016. And we take quite a broad view on what production means within that effort.
So you’ve basically been looking at policy from a higher level and tried to understand the global themes…
Yes, exactly, and the initiative now has more than 26 countries that account for, I think, something on the order of 80-85% of world manufacturing output. So the idea, the ethos here is that we try to get the views of multiple stakeholders, the people that look at this through different lenses – CEOs, ministers, union leaders, academics and so on – and typically try to incubate public-private efforts to solve the challenges of today and prepare the ground for the future.
What do you think are the challenges today? What are people trying to solve at the moment?
I would say the impact of technology in all its forms – there is an impact on business, there’s a societal impact, and then there’s also implications for the environment. So we try and take a view across those three main pillars.
We have projects on technology and innovation, one on what we call the future production workforce which deals with skills challenges looking ahead over the next few decades, and also accelerating sustainable production. How can we get production worldwide to align with UN Sustainable Development Goals?
So in your blog you were talking about production, and you have quite a specific definition for it, how would you explain it?
Well, as I said, we take a very broad view of what production is. A full value chain approach. So that includes all the way back to initial design, the sourcing of materials, then onto the shop floor, then through the supply chain all the way to the end user, and then there’s end of life issues to consider. So everything that comes under that umbrella is something that we regard as being relevant to production. And as such we have quite a diverse group of companies and organisations that engage with us.
And are there any global trends you’ve identified within that field, or production in general?
Yes, I think there are a few things we can say across those main pillars. I think what you find is that technology is heavily impacting business. I think it’s almost becoming an unavoidable, existential requirement for companies to embrace technology these days. And one of the challenges around that is capturing value from new technologies, including robotics, artificial intelligence, internet of things, additive manufacturing. How are these technologies coming together? Is there potential in using them in combination to obtain greater value capture? So these are the questions we are trying to tease out.
What you often find is that there are different sectors have embraced technologies at different rates and within each sector you will often have champion companies, companies that are sort of in the middle of the road, and then, for want of a better word, follower, laggard companies. And from a policy perspective you want to raise everybody up, as it were, especially SME’s who might not have the resources, whether it is human resources, or the capital to invest in and take advantage of these technologies. There’s a lot of support required to lift everybody up. So that’s one of the things we’re looking at.
Then there’s the impact on what’s happening not just on the shop floor but across the value chain, and what we’re observing from the employment point of view is that I don’t think we can necessarily predict the future in terms of whether there is going to be net job gains or losses, but I think but it is fair to say that some segments in the value chain will contract and others will expand. The eventual picture may not be a clear one, certain tasks will be fully automated, partially automated, etc., but overall it will entail requirement for other kinds of skills. So that’s not a settled picture and it’s one of the things we’re also very keen on investigating further.
They have said that technology has actually increased the amount of jobs than gotten rid of them…
Exactly. Every previous industrial revolution, going back 200 years, has in the long run created more employment and better employment. But then there is contemporary debate about whether this time it’s different.
Why would it be different?
That’s a very good question. I think maybe people feel like this revolution, these technologies that are coming on stream now, have the potential to make more people redundant. I don’t know, and nobody fully knows yet. In my work I encounter scepticism, pessimism, optimism, and every other kind of viewpoint, so it’s something we’re still trying to figure out.
So coming back to AI. They do say it will change the world more than the original industrial revolution, I mean I don’t know if you agree with this, but how do you think it will affect production?
I think it’s already affecting production significantly.
I think it’s probably the technology we have on our radar above all others. I think computing power’s reached a certain level that is enabling lots of other developments to emerge from it, and it’s getting more difficult to keep pace.
I think there is a sense of untapped potential at the moment, that there is so much data already being generated that companies don’t know how to use. The figure could be as low as one or two percent. So it’s almost as we’ve reached so far but we’re lacking in certain capabilities to fully capture the value.
I think there could be quite a profound change over the next fifteen to twenty years. And I think the companies that thrive will be those that learn how to take advantage and how to make sense of the data they generate, whether that’s on the shop floor or whether it’s in their consumer relations or whether it’s elsewhere in the value chain. The companies that get a handle on that will be the ones that will prosper more.
So I think it’s significant in the sense that I think it’s an unavoidable change, and it’s an unavoidable challenge for businesses large and small.
And which businesses, or which industries, do you think are currently leading this, or are most advanced?
Often when you’re dealing with a high value added product, I think automotive has embraced robotics quite a lot. I think their production lines are very sophisticated now and I think they tend to be at the forefront.
At the other end of the scale you might have a sector such as textiles, and we deal with countries that have significant textile industries. Here there can be an ambivalence about technology, because in these countries they see that this sector provides a lot of employment, maybe not very good employment, but it’s employment, and they’re sometimes reluctant to fully embrace technology for fear of creating a disgruntled population. And of course clothing is less sophisticated than a car. So you get this range of adoption. But I think the trend in all sectors is that it’s becoming more and more ubiquitous.
And how is this actually changing policy? I mean can you give as a concrete example of something that would change on the policy side?
I think on the policy side, going back to the issue of ASEAN, these countries fear the so called middle income trap. It’s one of these issues that development economists, who we work with, debate over. Is leap-frogging something that can be viably done? Can you jump from lower levels of development to higher levels of economic development by embracing technology? To do so will will likely involve huge upskilling.
Do we have any examples? Has that happened?
The classic example is mobile communications in some parts of the world that would traditionally have had limited landline communications technology, and in some cases never invested, because they jumped straight to mobile phones.
The challenge with this so called Fourth Industrial Revolution, and the amount of data generated in it, is that it requires significantly greater investment than mobile communication because the volume of data requires optical fibres, significant infrastructural investments, and so on. And that capital requirement is a challenge for countries that are trying to promote themselves to higher levels of development, and create an infrastructure and a platform for their companies to really take advantage of these new technologies. So that’s one of the debates that we’re having, and I think also on the issue of infrastructure, one of the other challenges is to think about not just the next five years but the next 50-100 years.
This comes back to the sustainability question as well, that the investment needs to be smart from that perspective as well. The work we do in sustainability is to do with demonstrating how technologies can be beneficial for supporting the environment, and we’re creating a policy toolkit to support this. Not descriptive as such, but informing the policy perspective to help meet the UN Sustainable Development Goals, and showing how technology in production can support that.
We have reviewed several sectors now, including food and beverage, textiles, electronics and automotive. And we have a further project focused on policy, what we term readiness for the future production and that is where we have analysed countries across many economic drivers, and looked at their institutions, and tried to determine how ready they are for the future of production. Again, not prescriptive, but a set of sort of policy tools that may be adotped according to the destination a country wants to reach, mindful that not every country aspires to being a high value added leading production nation.
At the moment we are creating what a so called Transition Framework that any country can look at and maybe think, okay, we’ve got an educated workforce but we don’t have a history of being a strong technological nation, what can we do if we wanted more high value manufacturing, what would we need to do, etc. etc.?
So basically we can use policy as a guideline or something to help us get somewhere rather than you know something to restrict us.
Exactly, that’s how we view it. We’ve already published white papers across each of these topics and they’re on our website. Those were published for this year’s Annual Meeting in January, and we’re already preparing publication of the latest findings for the next Annual Meeting in January.
Okay brilliant, that sounds all very exciting. So where do you see this all heading, I mean we’ve kind of been talking about it already but what’s the next big disruptor in your mind?
Well, that’s an interesting question. I think the question of sustainability in all its forms is the long term aspiration. So it’s about finding solutions and there’s no one silver bullet, but rather many actions required to sustainably live within planetary boundaries. But also sustainability from societal and economic standpoints, acknowledging the myriad contexts that exist in different countries and different sectors, that’s ultimately what we need to work towards. I think that involves strong levels of good employment.
Production, including manufacturing, is responsible for a lot of natural resource use, a lot of carbon emissions, etc., so there is sort of an onus to come up with a greener way of making things. From the economic standpoint, one of its stats that I perhaps should have mentioned earlier is that there is only at best about 30% of the companies that are really taking advantage of the technologies that are out there. A lot of companies are trapped in so called pilot purgatory where they try something as a pilot but never really go through with overhauling their operations, or else not try at all for whatever reason.
So it’s partly about trying to encourage greater embrace of technology, and I think largely speaking we are advocates, because it opens up the possibility of new product opportunities. That it allows us to tackleproblems that were not solvable before, while acknowledging there’s potential for unintended consequences as well, and trying to anticipate and mitigate those.
That’s a great way to think about it.
Returning to your question about the next big disruptor, I’m a bit reluctant to use the word disruptor, but I think there are incremental solutions out there, and some of them reveal considerable promise.
Over the last few years, we’ve tried to come up with what we call a Vision for the Future of Production, something that all of these different people that we work with can agree on.
We hear things that are kind of, I don’t know if the word is intimidating, but fairly daunting, such as that industry requirements are changing faster than the length of a college degree. How do you design education such that when you come out of college, you have an essential base of core skills, but then there’s also this notion that soft skills are going to become more important for technical people.
There are things like standardisation that are becoming more important, especially when you’ve got a lot of cross border data, so when you have that, when you work across jurisdictions, that’s where the soft skills come in when you have to work with a team from another company, from another country in another culture. For example, one of the challenges about additive manufacturing, given that you can produce a product anywhere, is that it brings up notions of IP and ownership. They can be a bit knotty to disentangle.
So standardisation is something that we’re probably going to become more involved in because we’re one of those organisations that do talk to people from around the world. Harmonisation of standards can be hard to get towards, it entails a lot of work and a lot of trust. One of the other things you also mentioned is this so called reshoring phenomenon, and it’s linked to this notion of labour arbitrage.
Labour arbitrage is to do with low wage countries that are often happy to stay as low wage countries, and I kind of talked a little bit about that earlier. Some experts think that there might only be a ten year window where they can continue to stay on that course.
You have this idea that for the consumer, products become more personalised and customised, and it might eventually become cost effective to reshore production in so called high value countries.
That is not an entirely settled consensus. Some of the people researching this think that we’re not really there yet, but I think it’s something that could happen. And that is a challenge for an emerging economy, you know, if they lose those external markets, what do they then do to lift themselves? So this is an example of the big policy conundrums for the years ahead.
Also, one of the interesting stats about production is that two third of R&D investment in a country is focused on manufacturing in one way or another. So the making of things is where most of research and development spend is directed.
Okay interesting, and that’s an average across the world?
Yes, and that’s repeated in many countries in various geographies at various stages of development. It’s one of the things that makes this topic so fascinating..
Selko is among the first 37 companies to join the Ethics in AI Initiative run by the Ministry of Economic Affairs and Employment, Finland. In the kick off event on Friday, the 5th of October, Selko represented deep-tech startups among multiple industry giants.
In this first-in-the-world initiative, companies working in the field of artificial intelligence, or using it internally, are challenged to discuss ethics around AI and to come up with a set of principles that encourage ethical use of data, the creation of algorithms and the implementation of models.
K-Group, Stora Enso and OP were leading the discussion on Friday and presented the work they had done internally. The conversation continued around political streamlining and issues around equality and working on project based AI.
As part of the initiative, Selko will soon be releasing its own set of principles and best practices.