How aerospace manufacturing embraces the smart factory.
How aerospace manufacturing embraces the smart factory.
Over the past decade, various technologies and systems have begun coalescing into one overarching theme within the aerospace manufacturing industry. This theme has been largely termed the fourth industrial revolution, or Industry 4.0. The fourth industrial revolution can be summed up simply as the integration of smart systems into manufacturing facilities to improve overall efficiency and increase profitability
Find out more in our report…
Download the Kingsbury guide to smart factories in aerospace
Research & design
At the beginning of any product development process is concept development, and then, research and design (R&D). This is especially the case within an aerospace manufacturing setting. As a critical but often time-consuming exercise, R&D will need to become more efficient to keep up with radical improvements in other areas. New technologies are being developed to assist engineers and designers with streamlining the R&D process to reduce the time and cost of researching and then designing new parts.
Generative design utilises machine learning to optimise components based on key parameters specified by the designer. These can include shape, volume, mass, structural strength, thermal dissipation, and even manufacturing technologies. Thousands of designs are generated by the algorithm, with each of these designs representing a unique solution to fit the design requirements. This dramatically reduces the time taken to develop new components and further helps improve the efficiency of the factory.
AR & VR
Augmented and virtual reality can drastically improve the efficiency of designers and engineers as it adds a 3rd dimension to the design process. This gives the designer more perspective of the problem at hand and therefore speeds up the iteration of concepts. Designs can be viewed in their true scale and engineers can discover issues that are not clear on a computer screen well before any prototypes are produced. After detailed design, engineers can mark up CAD models for updating by essentially viewing the virtual model of the assembly as if it were already manufactured. This is a highly intuitive process as it simulates the inspection of real-world manufactured parts.
Smart aerospace manufacturing
The core of any smart factory is the machinery used to make the actual parts. Existing manufacturing technologies are constantly being optimised and new technologies are being developed. The smart aerospace manufacturing facility of the future will consist of a range of highly specialised as well as multifunction machines that can work collaboratively to produce parts efficiently and cost-effectively.
Additive manufacturing (AM) has seen increased adoption within the aerospace manufacturing industry. It has transformed from a tool used primarily for prototyping to a tool that is capable of running production cycles on highly complex components.
AM is uniquely positioned to allow for multi-material manufacture, incorporation of embedded sensors for conditional monitoring, and light-weighting of parts.
AM technologies are continuously evolving in terms of speed, cost, and flexibility. No smart factories of the future will be able to compete without access to advanced AM machinery.
Augmented reality (AR) is already being used by various aerospace companies to assist in the assembly of complex systems. One key example is the assembly of the F35 raptor. AR works by placing images onto the work environment with the assistance of cameras and depth sensors that give the wearer a wealth of information. This information can include, for example, assembly instructions, quality control info and even wiring and pipe routing.
The amount of information available to the technician can dramatically improve overall efficiency and accuracy. Coupled with IIoT (Industrial Internet of Things), AR can display specific component details like what stage of the manufacturing process it is in, batch information, and even where it must be stored in inventory. The amount of information that can be made available to technicians on the floor is limited only by the imagination of the workshop managers.
Multifunction machines provide an aerospace manufacturing facility with extreme flexibility and lower capital investment. These can combine the best aspects of different stand-alone machines. An example of this would be the combination of an Additive Manufacture machine and a 5-axis CNC mill.
AM machines can create highly complex parts that are lighter and stronger than normal designs. After the part is printed it can be transferred into the milling bay where its surface finish can be cleaned up and features can be machined down to the required size, thus incorporating the strengths of both additive and subtractive manufacturing techniques.
In the aerospace manufacturing industry, tool management is a critical part of maintaining an efficient and validated manufacturing operation; this is due to the expensive tools required to machine exotic materials. If tool life is not monitored, the behaviour and lifespan of tooling is unknown, resulting in unplanned breakdowns which in turn introduce time wastage and increased costs.
Direct tool monitoring refers to the act of directly measuring the tool at predefined time intervals to gauge its current condition. This can be done with lasers to measure 3D profiles or microscopes to assess the condition of the cutting edge. This is a highly accurate method of measuring tool condition but has the disadvantage of interrupting manufacturing time.
Over time a database can be generated that can be used to reasonably predict the expected life of a tool, but these time frames are often conservative and thus result in more tooling being replaced than is necessary.
Indirect tool monitoring refers to the monitoring of tool life by measuring indirect signals such as cutting force, torque, spindle power, current, temperature and even acoustics. This method is less accurate than direct measuring techniques but has the advantage of running continuously during machining operations and with advances in artificial neural nets and machine learning, the system can become better at predicting remaining useful tool life.
Embedded sensors are a hybrid of direct and indirect monitoring. They are direct in the sense that the cutting tool is measured directly, and indirectly, so that machining operations are not disrupted. These sensors typically measure the resistance of the cutting tool as this is one of the easiest ways to determine if a tool is worn, chipped or broken.
These embedded sensors can be monitored by a machine operator or in the case of a smart factory, the monitoring of these tools can be handled automatically. An algorithm can determine the ideal time to change a tool to create the least disruption to productivity. The aerospace industry is renowned for manufacturing extremely
high-tolerance components that cannot afford to be damaged by dull or damaged tooling. Embedded sensors are the logical next step in tool-life monitoring.
Whenever smart factories are mentioned, the first image that often springs to mind is that of a factory filled with robots efficiently going about their tasks without any human input or interference. Industrial robots are taking on tedious, difficult and repetitive jobs in modern factories and cobots (collaborative robots) are assisting humans on the floor with more complex tasks.
Robots have typically been used in controlled or caged-off areas where the tasks they perform are predictable and highly controlled. Cobots are designed to work safely alongside humans in such a way as to react to unpredictable scenarios and to adapt to their surroundings with the end goal of improving productivity to higher levels than can be achieved by humans or robotics operating independently. This is especially beneficial in aerospace manufacturing where highly complex parts and assemblies are created that require robots to be capable of working efficiently with highly skilled technicians.
Cobots have become cheaper over the past few years due to healthy competition created by the void left by the larger industrial robot manufacturers underestimating the business case of cobots and will thus see increased implementation in the coming years.
Cobots are also being designed so that technicians on the floor can program them without needing to understand the complexities of how the system works. This frees engineers from having to interfere on the floor and lets technicians adapt robots on the fly when they see they are not performing optimally
Industrial robots will not be replaced by cobots but will do the heavy-lift, dumb jobs for which they were designed. One of the key uses of a robotic arm in an aerospace factory is pick and place operations. There are many examples of this; a machine can load raw material or preset fixtures into a machine and once the task is done it can remove the part and place it on a conveyor, where it can either be taken to storage or pass through a quality control procedure. It can even retrieve tooling from a centralised tooling bank and change it out with a worn or broken tool.
The future of aerospace manufacturing lies heavily in data management. The increased adoption of embedded sensors in manufacturing is creating a wealth of data that needs to be analysed for improved efficiency.
One of the more exciting aspects of the future factory is the growing inclusion of sensors within machines. As industry 4.0 methodologies take root in the manufacturing industry, the price of sensors and systems needed to enable the smart factory have become cheaper as demand has grown. In the past, embedded sensors were niche, expensive items that had an unclear return on investment. This was due to the lack of robust and adaptable software solutions that could make sense of and organise all the data collected into a useable form.
This is no longer the case. Data can be analysed to
illuminate potential areas of the production cycle that could be improved upon. Many machine tool manufacturers are including sensor systems in their equipment not only to assist the end-user but also to gather data that can be used to improve their machinery.
Raw data needs to be analysed to turn it into something useful. This can be done with machine learning (ML) algorithms that are capable of looking at a data set and finding patterns and indicating points of concern in the data.
Examples of machine learning algorithms
The slight increase in vibration of the spindle in a machine can indicate that the tool is about to break. Before any operator has noticed, the controlling software can order the tool be changed out with a fresh one. The data gathered on tool lifespans can then be presented to the workshop manager who can make strategic decisions that can improve the performance of the various subsystems.
Another example is a machine that can monitor itself and indicate to the central system that it requires maintenance. Jobs can then be distributed to other machines while a technician is alerted to the fact the machine requires maintenance. This preventative and even predictive style of maintenance ensures that machines last longer and that production is not unexpectedly interrupted.
Despite the fact that artificial general intelligence (AI) is still a long way off, it is not unrealistic to envision the aerospace manufacturing facility of the future making use of some derivatives of AI to manage all the different subsystems and make intelligent decisions with the end goal of improving efficiency.
As the economies of the traditional manufacturing hubs in Asia grow, so do the salaries of workers. China has reached the point where its manufacturing isn’t as cheap as it was 10 years ago and countries like Vietnam are leveraging their cheaper workforce to supercharge their economic growth as China did.
However, the writing is on the wall for cheap labour. Companies are realising that the better long-term play is rather to focus on automation than to rely on pools of cheap labour being available indefinitely. This change is evident in China, where demand for robotics has grown as the cost of labour has increased.
Furthermore, this has resulted in many companies bringing back manufacturing operations to their countries of origin. Robots do not care about geographic location, so companies can theoretically open up aerospace manufacturing facilities in any country around the world. This also allows countries that have been unable to compete within the manufacturing space to finally add manufacturing to their growth strategies.
It must be noted that while the rise of automation will reduce the requirements of cheap labour, there is still a need for highly specialised technicians that can operate, program and maintain these complex devices that cover a range of fields including software, electronic, industrial and mechanical engineering.
In short: the need for cheap labour will decline but the need for highly specialised and well-paid professionals will grow.
Lights-out manufacturing is the ultimate dream of any manufacturing company, especially within aerospace manufacturing facilities where labour is an expensive commodity. When it comes to efficiency, machines that work completely without supervision to manufacture and assemble components around the clock is the pinnacle.
Machines can operate 24/7 without rest and can inform operators in advance of any maintenance required. This means that other machines can take up the slack while maintenance is performed. A typical CNC machine can manufacture a complete part without human intervention.
However, the task of loading and unloading the machine was typically left for a human operator. This was one of the main hurdles for complete lights-out manufacturing as once a part was done the machine sat idle waiting for an operator.
With the rise of cheaper robotics, companies can now begin to retool their facilities to take full advantage of the cost savings and increased production capacity of lights-out operations.
As technologies and systems continue to evolve to drive efficiency in aerospace manufacturing facilities, the aerospace industry will be the first adopter of these cutting-edge technologies, as has always been the case. This is due to the fact that aerospace companies are always operating on the cutting edge and need to continuously innovate to stay competitive. The first truly smart factories will most likely be aerospace manufacturing factories.
A key point to note is that companies who do not stay abreast of these technologies will most likely suffer in the market as their profitability will inevitably be less than fully automated smart factories. The gains in efficiency and reduction of costs are simply an unbeatable combination. The initial capital investment will be high and there will undoubtedly be challenges such as lack of skilled labour and concerns over data security but the wave of automation in manufacturing is gaining momentum.
To find out how your company can begin gearing up for the next evolution in manufacturing, contact a Kingsbury specialist to assist your company in preparing for the rise of the smart factory.