ATMANIRBHAR:The buzzword today is open to interpretation by different people. For me it is the ownership of product IP. In the two contesting scenarios of designing and owning a product with lesser indigenous content (IC) and higher IC but no IP ownership, I always prefer the former. Premise is simple, if we own the IP, we know ‘why it works’. Once we have the system level understanding of the product, we can slowly indigenise the rest of the components. In short, it is not ‘Make in India’ it is ‘Made in India’. The difference is huge but bridgeable.
DESIGN OF FUTURE: THE DIGITAL TWIN
Digital Twin (DT) has many definitions. To put simply, DT is making an entire product inside a computer with all its functions defined with a set of mathematics. Change in one sub-system should ideally lead to a change in the performance of the system. DT should be able to predict this change. DT will have many facets like predictive simulations, visualisations, and data-oriented predictions. Predictive simulations would mean what is the effect of change in missile shape to its flight characteristics or change in engine location and/ or weight to the floatation of ICV.
One more important part will be the visualisation. A ship would look like a real ship on the computer and user will be able to walk into the ship and ‘see’ various systems in operation. User reverses the throttle and the ship slowly turns back. Mind you, it is not animation which is a small part of DT visualisation. DT will ensure that the speeds predicted, forces generated, penetration of projectiles, working of mechanisms are correctly modelled.
The third leg of the DT will be the acquisition of data from existing system and converting that raw data into knowledge. The know-how will then be used to make a DT for that system. This DT will be dependent on the learning from the working of the system. The focus of this article is mainly on the predictive part of DT.
IS IT REALLY NEW TECHNOLOGY?
A Yes and a No. In some form or the other DT has been existing since the advent of computer. As early as 1965 scientists predicted the heating of a launch vehicle at supersonic speeds, loads on the airplane wing while flying or effect of wind on trajectory of artillery shell. These methods were accurate enough and were limited not by the maths but the computing power available. The idea and its mathematics have been understood for quite some time without the hardware available to implement it.
Coming to the No part. Internet of Things (IoT) techniques that are integral to development of DT did not exist in those days. Artificial Intelligence (AI) and Machine Learning (ML) algorithms are new. These integrated with predictive algorithms will lead to newer designs with far more accuracy than was previously possible. The tool would suddenly revolutionise the way we design new products. Collaboration with multiple teams will become real time as all of them will contribute their sub-system models that will be connected to each other by the chief designer. Separate teams with expertise in visualisation and animation will make the entry of person look real inside the system.
ARE WE LATE? WHAT TO DO?
I think I have sufficiently established that DT will be significant part of our future. It is not that the world is very far away from us. Having put concentrated effort we can bridge the gap that we currently have with the western world. The deciding factor is that we do not need massive investment of testing infrastructure, labs etc for making a DT.
The ingredients are simple – a large enough computer and manpower trained in mathematics, physics, and code development. CDAC has been successful in making high performance computers having capacity more than what is required by the country and it can be safely said that we will not be found wanting in this area. There are other government organisations, which have computing power rivalling that of CDAC. In all situations if not unfavourable here.
Good code developers are available in abundance in India. These same people, however, lack the relevant knowledge of physics and maths. If DT is made a priority with requisite funding, I am sure we will be able to attract talent that is otherwise going in banking, finance and IT. The sense of achievement when the DT works for a complex defence system cannot be explained.
To reiterate we are little late here but not behind others. To test this technology, one or two medium size defence programme will have to be fully designed using DT approach. Like Dr Kalam envisioned about fluid dynamics in 80s by setting up Centres of Excellence (CoEs) at various places, DT will require same commitment. Two successful programs and DT will catch the imagination of top management and job seekers alike ensuring steady support from the top and continuous flow of manpower from the bottom. To show by example, please see the firing of FSAPDS on an enemy tank. Graphs of flight are plotted below to show it is not a blind animation but a scientific work.
CONCLUSION
Though we missed a lot of buses in creating defence and testing infrastructure, we cannot miss this time. A software hub of the world needs to produce its own software. Draft Defence Acquisition Procedure 2020 (DAP 2020) is supportive now with the addition of a draft chapter on procurement of ICT and software systems. The need is to identify two-three ambitious projects and start designing them the DT way. Once successful, the usefulness of DT will be demonstrated in a functioning product that is “Made in India”. Just be careful not to be wrong with the Maths and Physics, it can shoot you in the foot!
-The writer is Vice President, Strategic Partnership, Zeus Numerix Pvt Ltd. He was one of the 10 innovators to meet US Defense Secretary, and is also a Winner of Lockheed Martin Innovation Award and various other awards. The views expressed are personal and do not necessarily reflect the views of Raksha Anirveda
