A Career in Three Acts: Prith Banerjee: As Chief Technology Officer for ANSYS, Prith Banerjee oversees simulation-based product design.
Phase two of my career was working with startups. I founded two small software startups based on technologies developed at the university. For both those startups, I was able to raise money from venture capitalists and build a product, and I was able to sell both of the companies.
For the last 11 years, the third phase of my career, I've been working in the corporate world. I spent about five years heading up HP Labs and as CTO of two large industrial companies, ABB and Schneider Electric. Last year, I joined ANSYS as its Chief Technology Officer. Things come full circle; when I was in academia, I worked in electronic design automation, high-performance computing, parallel algorithms, and so on; today, my job is again working on simulation tools, but in a broader way. The tools are not just electronic, but electronic, fluid dynamic, and mechanical.
In my career, I've seen big changes in the way industrial companies develop engineered products. The way you made engineered products in the past was to design a product, develop a prototype, and test it. You might end up building a hundred prototypes and doing crash simulations or other kinds of destructive testing on them. This takes a long time, because you develop these prototypes manually.
In today's world, companies use simulation-based product design. They still use a CAD tool to design, but they also use it to simulate the tests. Until recently, the simulation models were not accurate enough; companies still had to build prototypes and the infrastructure to test them. You might have to build a wind tunnel that could simulate wind speeds of 300 miles per hour to test an airplane model, for example; the wind tunnel test would tell you whether that wing was actually developing the projected lift. These days, you do wind tunnel simulation through computational fluid dynamics. You don't need a wind tunnel; you don't have to build that physical prototype.
In the last five years or so, industry has moved from single-point designs, where the focus was on fluid dynamics or mechanics or electromagnetics, to multi-physics simulation. For example, imagine a 3G wireless antenna on top of a beam structure somewhere in Florida, with Hurricane Maria coming with winds of 200 miles per hour. The beam structure supporting the antenna starts bending, and the antenna structure, which was circular, becomes oval, and because it is oval, the electromagnetic wave patterns get distorted and you don't get the full 3G connection. Structural mechanics, fluid dynamics, and electromagnetics all interact. Tools and computational capabilities have advanced so much that you can simulate all these things with high accuracy; you don't really need prototypes. That's a sea change in product development.
The next big trend that I have seen for industrial companies is connected intelligence, the Internet of Things. The first use case of these connected products is to let the owner know when the product has failed so appropriate services can be brought to bear. The next use case will be to use that data, together with simulation tools, to predict a failure. We talk today about "pervasive simulation," where simulation is occurring at all phases of engineered product design.
The ability to do this quickly can expand the design space in which you operate. It can enable you to design products that you could not previously design. Add AI and machine learning, and you can begin to do generative design. With generative design, you give the simulation tool some top-level boundary conditions, and the tool automatically examines 10,000 different cases and presents the top three optimal designs. It's amazing how digital technology has enabled industry to transform in the last 25 years.
Published by Taylor & Francis. All rights reserved.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||INNOVATION C-SCAPE|
|Date:||Nov 1, 2019|
|Previous Article:||Beyond Performance 2.0: A Proven Approach to Leading Large-Scale Change.|