Why mechanical engineers should learn A.I.




Chair designed by designer Philippe Starck with help of Autdesk generative design


According to a survey by Gartner, number of enterprises using Artificial Intelligence (A.I.) have increased by 270%, creating a shortage of AI professionals.

[1] AI is going to affect every profession, but how will mechanical engineering get along in this future scenario? There are some mechanical engineering fields in which AI is about to give a paradigm shift.

AI Helping in Complex CAD

AI used in Computer-Aided Design (CAD) generally works on knowledge-based systems. Design artefacts, rules, and problems in CAD are stored which later assist CAD designers. Merging of AI and CAD is done through Model-Based Reasoning (MBR). Many new releases of software packages are using knowledge-based systems. A major field for the application of AI is Generative Design. Generative design tool takes design requirements as input and gives possible designs as output. SolidWorks gives a feature of topology optimization in its 2018 version by using different algorithm based on generative design.

Autodesk launched a project named Dreamcatcher which offers the feature of generative design. Using this utility, instead of designing by the hit-and-trail method, engineers can select a design provided by software after observing suitable trade-offs for any features.

Artificial Neural Networks in CFD

Computational Fluid Dynamics has been of great interest among scientists, engineers and mathematicians. The turbulence and chaos associated with fluid mechanics have made it a lot difficult to solve with Direct Numerical Simulation (DNS). There are some models available, namely Reynold’s-Averaged Navier-Stokes equation (RANS) and Large Eddy Simulation (LES), which approximates flow behaviour and AI also found its way among them. Artificial Neural Networks(ANN) are gaining interest in academia for their potential to give approximations of flow with less computing power, time and dimensional reduction of problems. They are also showing good agreement with traditional CFD models. The challenge is to train ANN with lots of example simulations. Also, you can’t get an insight of flow mechanism with neural networks.

IoT and Data Analysis

4th industrial revolution is going to connect all machinery in a production plant and consumer products, so engineers can analyse, optimize and ensure quality of the product. Managing such technical data will require engineers who could read between the lines of sensor data. Mechanical engineers with AI skills would be required to work on software which can handle data provided by sensors in components of power plant, production facility or consumer products. One example of data science use in power plant optimization. Data collected from Supervisory Control And Data Acquisition (SCADA) can help predict failures, avoiding any loss of money or life.

A US-based company Sparkcognition is providing solutions to power companies that detect anomalies in plant data and predict any failure sufficient time ahead, avoiding downtime and loss of revenue. [3] AI is creating strides in self-driving cars as well as industrial robotics.

Early fault diagnosis framework using AI techniques for rotating machinery systems

How to Prepare for this Upcoming Trend

To prepare for this future scenario, one should start working after some research and planning. I asked some questions from Mr. Ricardo Garcia Rosas, a Ph.D. scholar in robotics and automation at University of Melbourne, for some insight into AI in mechanical engineering:

What specific languages and skills should be focused on while learning AI, with mechanical engineering in view?

“In terms of programming languages, Python is the most widely used in machine learning (ML). For data analysis, both Python and R are good options. My view is that most mechanical engineers will use AI/ML as part of a tool, e.g. CAD/CAM, FEA software, or to support data analysis and decision making.”

Which mechanical engineering fields can have breakthroughs by AI?

“I think that areas that involve design and analysis will be the first to see the benefits of AI/ML, and are already being introduced commercially. Areas that require some form of optimization, estimation and/or evaluation of potential solutions have the potential to be disrupted by AI/ML. Some of the work I’ve seen is in areas like material synthesis, generative design, and mechanical fault estimation.”

What kinds of tasks can AI take over from mechanical engineers?

“I think it’ll be similar to what happened with factory automation but from a computational perspective rather than physical. In factory automation, the repetitive and strenuous physical tasks were taken over by robotic manipulators which enabled people to focus on more dexterous tasks. With AI/ML a lot of the data wrangling and computation is going to be automated, mechanical engineers are then going to be able to focus on using their analytical skills and expertise to make decisions.”

As a conclusion, one can understand that all knowledge and creativity based work can be disrupted by A.I. Much of our problems can be solved by ingenious designs made with its assistance. With further advancements in A.I. itself, more of its applications can be found.

As CEO Alphabet Inc. and Google Sundar Pichai stated:

“AI is probably the most important thing humanity has ever worked on. I think of it as something more profound than electricity or fire.”

References

[1]

Gartner, “Gartner Survey Shows 37 Percent of Organizations Have Implemented AI in Some Form,” 21 January 2019. [Online]. Available: https://www.gartner.com/en/newsroom/press-releases/2019-01-21-gartner-survey-shows-37-percent-of-organizations-have.

[2]

R. Deplazes, “Philippe Starck Partners With Intelligent Generative Design to Co-Create His Next Masterpiece for Kartell,” 19 April 2019. [Online]. Available: https://adsknews.autodesk.com/news/starck-intelligent-generative-design.

[3]

Sparkcognition, “Predictive Maintenance for Fossil Power Producers,” 26 October 2020. [Online]. Available: https://www.sparkcognition.com/predictive-maintenance-for-fossil-power-producers/.

[4]

Y. Wei, Y. Li, M. Xu and W. Huang, “A Review of Early Fault Diagnosis Approaches and Their Applications in Rotating Machinery,” entropy, p. 26, 2019.

 

Next Post

Salesforce (CRM) Expands AWS Partnership to Enhance App Development

Salesforce CRM recently announced that it is expanding its partnership with Amazon’s AMZN cloud arm, Amazon Web Services (AWS), to make their combined offerings more accessible to customers. This extended partnership will help customers rapidly develop compelling applications to accelerate the digital migration of their businesses. Software developers will be […]