Page 38 - AAA NOVEMBER - DECEMBER 2017 Online Magazine

COLUMN NAGENDRA VYAMAJALA

Seismic Shift

Smart solutions using Artificial

Intelligence (Machine Learning)

rtificial intelligence (AI) is making and also progressed significantly in AI
rapid strides in all walks of digital initiatives to its top global aircraft
A life. While science fiction often manufacturing customers.
portrays AI as robots with human-like One of the world leaders in aircraft
characteristics, AI can encompass industry approached us with a problem
anything from Google’s search algorithms in design engineering that was taking a
to IBM’s Watson to autonomous weapons. lot of time to provide a design solution
There are literally unlimited opportunities with the existing established process. was faced with the situation of more time
to apply this technique in the real world Tech Mahindra with its team of domain in quality control in their shop floor. The
with a multitude of benefits to derive from. specialists and AI experts recommended current job involves a quality inspector
All industry verticals be it automotive, a solution to this complex design problem to go through 100 percent sampling
transportation, medical, oil and gas, using Machine Learning. The amazing checks (in this case it is mandatory) of the
energy sector have realised the potential performance of the Machine Learning manufactured part. This has an impact on
of AI in the recent past and have now algorithm based on a supervised solution the productivity, hence Machine Learning
started embracing it. There are multiple using platform agnostic software, techniques were deployed for this. The
forms, subsets of AI that can be suited and provided outstanding results to this algorithm scans the inspection quality
adapted to a variety of problems to get otherwise highly complex problem. It reports of the part and prepares a signature
the right set of benefits. analyses huge data points of the previous corresponding to it. At the first level, it
The aerospace domain is mainly driven similar situations and provides the closest reads the scans and decides whether the
by the need for quality and processes solution mimicking human knowledge and quality is good and the part can be passed.
which are tested, qualified and certified awareness. The algorithm learns on the In case the first level scan doesn’t lead
over decades. Any new technology that data it has and equips itself for the job. At to a decision, it refers to more detailed
needs to be implemented in this domain the same time if the situation is unknown, scans of the part till the robust algorithm
has to be well matured. AI which has shown the algorithm indicates the need to go for converges on the size of the defect. The
a prevalent and dominating presence in manual intervention. The results tested accuracy of the AI tool in determining the
other engineering domains has made its for a variety of conditions are robust and defect and extent of defect is close to 98
positive impact even in aerospace. Almost encouraging. Tech Mahindra is now gearing percent. Many more features are being
all top global aircraft manufacturers have up for an increased scope of deployment added to make this AI tool more powerful.
started seriously thinking to implement of this tool within the customer network. In future, AI is going to be a strong
AI in any possible way in their industry, It helps the customer in reducing the cycle digital technique that will change the face
be it in shop floor, design engineering, time by almost 30 percent and improving of engineering in all industry verticals
quality etc., Tech Mahindra which has the quality by more than 50 percent. through its strong machine learning
already established digital initiatives Another North American aircraft algorithms that benefits the customer
like IoT, robotics, Augmented Reality component manufacturer and supplier in cutting down the cost, cycle time and
and Virtual Reality (AR/VR), Industry 4.0 of parts to the aircraft industry worldwide improving productivity and quality. The
beauty of AI is that it can be coupled with
All industry verticals be it automotive, other powerful digital solutions like AR
and VR, Big Data and Industry 4.0 for even
transportation, medical, oil and gas, energy greater benefit.

sector have realised the potential of AI in the
recent past and have now started embracing it. Nagendra Vyamajala, Aerospace
Engineering Practice, Tech Mahindra
38 MAINTENANCE REP AIR O VERHA UL OC T WWW.GBP.COM.SG/AAA
.GBP
.COM.SG
WWW
38 ASIAN AIRLINES & AEROSPACE NOVEMBER / DECEMBER 2017OBER / DECEMBER 2017