Software company Cimteq can help make a cable factory smarter
The complexity of making a cable brings to mind the old rhetorical question about how long is a piece of string. An appropriate image, as the length of a cable is one of the factors that make cable manufacturing difficult, although there is more to it than that.
Short simple cables, think phone chargers, can be easily mass-produced. Their design is a number of strands of metal conductor, normally copper or aluminium, twisted together and covered by plastic insulation. Once designed, mass production is trivial, and the customer simply selects from a product menu.
However, long complex cables used in oil and gas projects come with many more challenges. Subsea cables are exposed to corrosive salt water. They suffer violent impacts. Downhole cables have to operate reliably at high temperatures.
While cables find their way into just about every operation across the industry, the largest and most complicated are subsea umbilicals. These contain multiple different conductors of power, data, hydraulics, injection gas, and can stretch for thousands of metres. To prevent signal contamination conductors need to be insulated against each other, as well as armoured to protect against their harsh operating environment.
That makes them expensive. Lifespans are partly a function of layout – conductors can wear against each other. Poor layout makes the cable less flexible, also shortening life. Good cable design is, in short, a key input to maximising life and minimising depreciation.
Cimteq, a company based in Wrexham, UK, has developed a software suite to support the design and manufacture of cables. The company’s design tool – CableBuilder – helps the designer specify the detailed construction of a cable, and then anticipate the manufacturing issues to be solved by reflecting the actual manufacturing conditions on the shop floor.
It models the real world through a user-configurable set of machine and production rules. Doing this means that design and manufacturing both sit in the real world, not a theoretical perfect one. This in turn means less manufacturing wastage, quicker production runs and better quality assurance.
When it comes to the actual cable build, a key problem to manage is length. If a batch of a thousand pieces of a one-metre cable contains 1% that are too short, you lose 10 items, a small investment that can be made up quickly in the next batch.
If a single 1,000-metre cable is 1% short, the entire cable may be wasted. Overspecifying length may be an expensive way of covering that risk, and also, if the cable is designed to a specific length and task it has to be exactly the right size. Overspecification also generates waste and scrappage.
Equally, in a 1,000-cable batch you can discard failed cables, but if there is a manufacturing fault along a single 1,000-metre cable you have a much larger problem. Just locating where the fault lies in the cable is a challenge. In the event of a repair, it is difficult to correct the fault without weakening the cable or leaving a join, which might critically alter the outer diameter. In the worst case scenario, a fault might render the whole cable a failure.
Every process, element and metre contains a chance of error. The key data capture point in manufacture is length. Length counters are commonplace, and detect stops in production and communicate with operators to initiate repairs. They can also schedule an inspection once a set length of cable has been run. The next step for smart length counters is to alert other machines in the process to stop or slow down in sync. They can finally provide data on how an order is progressing.
Cimteq steps in here with its Cable Manufacturing Execution System – CableMES – which collates and collects length data from a user-designated set of machines (using the Historian tool to process data) and presents this to the operator on a tailored dashboard. This allows the cable manufacturer to hit exact lengths, and also combine data capture from test equipment live, to identify quality issues quickly.
While some materials used to make cables may be similar to one another, they function very differently. The supplier’s product codes can only differ by a single digit and human error can easily cause the wrong material to be used.
By logging the materials, CableMES’s Scan Materials feature provides an early warning of a difference between the bill of materials and raw materials. Catching this early can save tens of thousands of dollars in wasted time and material.
Cimteq believes that its suite of products is a central part of the expansion of the concept of smart factories into cable manufacture, which has historically been slow to adopt IT as a core component.
While basic sensors are becoming commonplace, it is rare to find a cable plant in which the data they produce is integrated and processed intelligently. The challenge is not one of computing power – while there are a lot of sensors, data processing can be effectively carried out on basic IT architecture. So, integrating Cimteq’s solutions into a medium-sized plant comes in at around US$100,000. Most of Cimteq’s work is retrofitting, which avoids disrupting production.
A length counter is the most versatile sensor for cable manufacturing and provides the minimum number of sensors to go into a smart factory. However, there other sensors that can provide usable information if it is properly communicated.
Temperature and pressure gauge readings can be collated with cable quality to find savings in materials and prevent scrappage. An automatic diameter measurement system can spot deviations in the dimensions of the cable, which may reveal an internal flaw. Scales too can reveal how a fault can be related to material use. Together, these can help speed along the diagnostic process.
Combining the pieces
The true genius of the smart factory is how it combines data from multiple sensors. Take a vibration sensor – this can detect abnormal vibrations with more sensitivity than a human being ever could. Vibrations are both a sign and a cause of a larger malfunction.
While on its own a vibration sensor can be used to diagnose and pre-empt a potential fault, by combing its data with data from a length counter and timing system, the length of cable made while the machine was vibrating can be tracked down and checked to ensure it was not damaged while the machine was malfunctioning.
In the end, data is only as good as the actions it can dictate. By using the data to control hardware and enabling people to make quality decisions, the smart factory comes to life.
A digital twin can aggregate the information from smart assets and product design knowledge to make a real time model of the factory. Not only can this help manufacturers organise the factory, it can form the basis of intelligent automated decisions. Digital twins can be written in a language more easily understood by a computer, so smart assets can better understand one another.
Automation is another distinguishing factor of a smart factory. While it is arguable that automation has been the basis of a factory since the Industrial Revolution, a smart factory combines automation with machine communication.
Robots, like the autonomous mobile robots visible in every report on Amazon’s warehouses, are an example of how advanced automation has become. These can be added to a factory without having to alter its layout and free up more valuable human staff from the dull work of moving components.
These can take orders from connected assets to bring raw materials to a machine before it runs out, or remove finished products.
3D printing has a place in a smart factory. Forget the crude plastics coming out of cheap consumer printers – complex and high quality metal and electronic parts can be made to order. A diagnostic system in a machine can communicate with an on- or offsite 3D printer to create a replacement part, reducing downtime. A smarter and better connected AI can look up a stock database to see if a spare part is available or make decisions on how urgent a repair is. With advanced enough predictions, replacements can be made or ordered before a part wears out.
The cable market was hit by a decline in demand around the turn of the century. However, the global energy transition will drive demand for energy cables by an estimated 5% per year until 2023. Investments in the renewable energy sector will drive investment in high-voltage cables, while electric vehicle growth, particularly in China and India, will see low-voltage cable demand grow.
Cable makers can use this growth to optimise their operations. As the factory evolves, it will be the smartest that thrive.