So if you think forward, now you could have your certified combat system on the ship, through operational test, ready to go fight. You could have your next system in test on the ship, being tested as you're operating in the fleet. You could have your third system with a bunch of algorithms doing AI kind of things, all on one single ship, and that can fundamentally change how we approach things.
We took a virtualised complete combat system on an old destroyer, hooked it to the sensors, shot a missile and hit a target. And so, we think this is going to be a pathway not only to really revolutionize how fast we can modernise our current systems, but then keep up with things.
“If you can’t solve a problem, make the problem harder.” If we made the problem so hard that we got only a week to update systems, not a year, then I think it will drive us to some of these new concepts.
Navy thinks this Digital Twin effort could help it get to a unified combat system more quickly. Some of the pushback on that effort is that if there is a vulnerability on one ship, a unified combat system means there is a problem on every ship? How do you make sure you are keeping ahead of that threat?
It’s something we've got to address, stem-to-stern, through the enterprise, and really bring security back as one of our fundamental things we think about when we put acquisition strategies together.
Having said that, getting the digital twin being able to upload 24 hours from compiling the combat, all give you a lot more resiliency, because you can update your systems faster as you find vulnerabilities, you can deal with them faster.
If you can have three versions of the combat system on the ship, and one gets corrupted, we can immediately deal with that. So, it’s actually a means of resiliency, not as a vulnerability. We’ve got to think our way through it.
The other thing we're seeing digital twin on more than just combat systems, we're now building digital twins of our shipyards. So how do we improve the efficiency of how we repair ships? And looking at, “Okay, let's look at the whole shipyard. Let's model it, let's simulate it, let's figure out where there's efficiencies.” So you're going to see that continue to kind of cross all the different domains in the Navy.
Consider an aircraft mission space where a critical spindle is about to snap. Without a Digital Twin Builder, the whirring machine will give no warning of its impending malfunction, and its failure will come at a random moment.
Diagnosis and repair will be relatively slow, constrained by data collection and the organising of human and material resources. For a fast-moving military mission, the lag can be costly. Because of the high value, we're talking about a lot of money for every hour of down time.
Problems can be mitigated with a Digital Twin Builder enables machines to have sensors so that the spindle in the machine is continually monitored and its performance data sent to a control room where the data is fed into a computer that acts as a digital twin of the machine – a virtual copy that accurately reflects the machine's current operating status based on real-time sensor data and physics modeling.
The Digital Twin, detecting a slight wobble in the spindle, might adjust its physical counterpart's operating parameters to correct for the wobble. Or, if the wobble can't be corrected for, it might warn of an impending malfunction.
In the control room an operator would then be tipped off, and set off a standardised response. "The Digital Twin creates maintenance work order, determine whether you could reroute production so that delivery is not impacted, find the appropriate maintenance personnel, and tell them where to go.
Maintenance personnel could quickly repair or replace the faulty machine, guided, either by a tablet showing a customised data feed on the impacted machine, or by a set of augmented reality glasses. Digital Twin technologies like this enables workers to minimise down time while improving performance.
"If some maintenance or repair task is beyond the skill set of a worker, using our tools, someone with the requisite skills somewhere else could see exactly what the worker standing at the machine sees and guide them on what to do.
Digital Twin robots have been designed to physically assist a human operator with tasks such as moving hot or heavy objects. At the moment, information flows only from the physical robot to its digital copy, but Digital Twins will eventually be bi-directional, so the digital version can adjust the operation of its physical twin in real time, instantly reacting to the information it receives.
Typically, by the time a customer delivers feedback about product quality problems, it can be difficult to diagnose what might have caused the problem so we can go back to a particular day, look at all the dashboards, drill down, diagnose and solve the problem."
"The goal is to assist workers by aggregating and crunching data, then creating insights from it. The control room will provide information and predictions, but the human has to make the decision."
But eventually, AI will take on a greater share of the decision-making burden. "As the systems become more intelligent, we can move to full automation by AI. That can relieve humans to focus on things that AI can't do: relationships, supply chain or customer issues, and managing workers."
So data will flow not just inside individual smart factories, but also between them. "Machines will be able to talk to machines directly, including machines located outside the factory. Networked machines will work together to predict failures, and to respond to them. For example, "the machine in your factory could tell a machine in an offsite factory shut down, we don't need the part you're making.
Many organisations are sometimes reluctant to test the technology on their own because they didn't want to disrupt their own facilities. It's often difficult to bring together experts in manufacturing with the wide range of analytical scientists that could provide useful input.
Digital Twin Builders provides industry with the freedom to experiment and build, then to take back new technologies and best practices to field level missions.
"The merging of physical and digital worlds disruptively affects the way in which products are manufactured, placed on the market and operated. By utilising the insights produced by digital twins, users will be well positioned to exploit the breakthrough this technology brings.
New solutions can not only simulate products and facilities during the product development phases, but also in manufacturing and—more importantly even when the products or facilities are in the hands of end-users.
As an example, by simulating the operation of a field formation, based on digital twin solutions, operators can test which flows in the most complex structure are most efficient to run under specific conditions. Through this type of simulation, it is possible to iterate and select a line of operational benefits.
Simulations use the enormous amounts of data generated by sensors in the assets, and let agents gain valuable insights that can improve a process and provide a basis for improving future similar processes.
It’s now possible to develop hybrid models that leverage machine learning with multi-physical simulation models to accurately predict why a process in a facility may fail after it has been implemented.
New technology covers a product or plant's digital twin to simulate behavior in different environments and stresses, the system is intended to predict problems before they occur. The prediction is based on information from physical sensors and physics-based analysis based on simulation models to provide results in 3D visualisation.
"The synthesis of the digital and physical asset will enable companies to capture value throughout their product lifecycle. “This solution helps equipment operators and service providers predict and improve asset performance and reliability with technical insights. A digital twin that merges technical models, manufacturing details and operational insights is unique in the industry.”
Enabler technologies like AI, big data and predictions have found their way into the real-world mission space and manufacturing domains. Existing Operational/Info Tech systems are not designed to cope with the masses of data generated by fully connected shop-floor applications.
High-performance computing is transformative, centered on the idea of high volume and volatile streams of data and massive compute power to perform analytics, and AI applications on top of the data, while existing systems are locally optimised technology.
Large industrial vendors have pushed the concept of the digital twin— a full integration of the physical with the virtual world. In essence, this means that all product design data is available at the time of production.
As an example, the full design data of a car body is compared in real-time with the as-is built data in a car body shop. In addition, production-relevant data is fed back into the design process, also reflected as closed-loop engineering. This frontloading of information will allow better decisions at a very early point of the product lifecycle and generate additional value.
With asset data available in real-time from production through data aggregation and AI predictive and prescriptive applications will generate insight to increase overall equipment efficiency and lift value from manufacturing assets in operation.
But what happens as soon as insight is generated and the status of the physical process needs to be changed to a better state? In manufacturing for discrete and process industries, the process is defined by fixed code routines and programmable parameters. It has its own world of control code languages and standards to define the behaviour of controllers, robot arms, sensors and actuators of all kinds.
Stable for decades, Control code resides on a controller and special tools, as well as highly skilled automation engineers, who define the behavior of a specific production system. Changing the state of an existing and running production system changes the programs and parameters required to physically access the automation equipment—operational equipment needs to be re-programmed, often on every single component locally.
To give a concrete example, let’s assume we can determine from field data, using applied machine learning that a behavior of a robotic handling process needs to be adapted. In the existing world, production needs to stop. A skilled engineer needs to physically re-teach or flash the robot controller. The new movement needs to be tested individually and in context of the adjacent production components. Finally, production can start again. This process can take minutes to hours depending on the complexity of the production system.
Current production systems are trimmed to high stability and low variability. For example, automotive production has a large number of product variants, but still has an inflexible production system. With a car model lifetime lasting several years, production managers have learned to live with this inflexibility and value stable processes. However, customer- and technology-invoked trends will increase the need of fully flexible industrial control systems.
First, the speed of innovation increases due to improved design systems and customer demand. Second, requirements for customisation increase steadily. Third, intelligent algorithms will produce a steady stream of proposals for process improvement.
If we assume that AI will fulfill its promises, the majority of manufacturers will be able to gain constant insight. Companies that are able to execute these insights faster will have a competitive advantage. Additionally, every new state of a production and supply chain system will be seen as a new experiment and fed back into AI systems, ultimately generating a cycle of a self-improving system.
All control code has a digital twin in the virtual world. The local instance is constantly updated from the virtual master code. A virtual model of the whole production system provides context for control code—these manufacturing planning systems are already heavily used by e.g. line builders for automotive providers.
Having a full digital twin of a production system, including control code, insights can immediately be pushed down to change the state of a production system.
Still, human interaction will be required. Production systems will optimise themselves based on simulated and real experiment. Improvements will rapidly be propagated and labor will optimise the learning, not the system. This could also differ over time or by external influence.
With the release of edge platforms, the technology is here today to minimise the time from insight to reaction. The power of AI is and will be the core enabler to realise automation fast and at an affordable cost.
1. Field level maintenance is generally characterised by on-near system maintenance, often utilising line replaceable units & component replacement using tools and test equipment found in the field-level organisation not limited to simply "remove and replace" actions but also allows for repair of components or end items on-near system.
2. Field-level maintenance includes adjustment, alignment, service, applying approved field-level work orders, fault/failure diagnoses, battle damage assessment, repair, and recovery to always repair and return to the user include maintenance actions able to be performed by operators.
3. Crew maintenance is responsibility of using organisation formally trained operators/crews from proponent on specific system to perform maintenance on its assigned equipment, tasks consist of inspecting, servicing, lubricating, adjusting, replacing minor components and assemblies as authorised by allocation chart using basic issue items and onboard spares.
4. Operator/maintainer system specialists for example, signal, military intelligence, or a manoeuvre unit receive functional individulised training from proponent on diagnosing/troubleshoot problems focus on system performance/ integrity identify, isolate and trace problems to on-board spares deficits correct crew training deficiencies.
5. Maintainer work orders accomplished on a component, accessory, assembly, subassembly, plugin unit, or other portion either on system or after it is removed by trained maintainer remove and replace authority indicates complete repair is possible return items to user after work order performed at this level.
6. Sustainment-level maintenance generally characterised by “off system” component repair or end item repair and return to the supply system, or by exception, back to the owning unit performed by activity function to be employed at any point in integrated logistics chain.
7. Sustainment level intent to perform commodity-oriented repairs on all supported items return to standard providing consistent/measure level of reliability execute maintenance actions support force and supply system not able to be performed at field-level maintenance unit.
8. Exceptions made to when in-house sustainment level maintenance activities may conduct maintenance and return items to using unit but also may be performed by contract agreement comprised of below depot sustainment.
9. Below depot sustainment level maintenance assign to component, accessory, assembly, subassembly, plug-in unit, or other portion generally after it is removed from system. The remove and replace authority indicates complete repair is possible at below depot level return items to supply system also applies to end item repair and return to the supply system.
10. Depot level maintenance accomplished on end items or component, accessory, assembly, subassembly, plug-in unit, either on the system or after it is removed define remove and replace authority indicates complete repair is possible at depot level return items to supply system, or by exception directly to using unit after maintenance is performed