Past Grants 

1. Title: Advanced Energy Management and Data fusion for Smart Homes Grids/ Heating and Cooling and Predictive maintenance using AI. Grantor: Liteon Singapore, Amount $2.06 Million Dollars. 2024. (Technical Disclosure/ Patent filed). 

2. Title: Evaluation of Piezo Sensors for effective use in Health Monitoring of Engineering Structures. Young Scientist, Fast track scheme.  Nov 2011 - Nov 2014, [Grant No: SERC/ ET -0324/2010], Dept.  Science & Technology, Ministry of Science & Technology, New Delhi, India, US $ 46,000, (INR 20.50 Lakhs), (Grant turned down as I moved from India to Singapore to expand my objectives to include components of mechanical structures with the help of Assoc. Prof. Pang Hock Lye, John, School of Mechanical & Aerospace Engg, NTU, and filed a Singapore Patent). 

Details of Filed Singapore Patent Title: Electromechanical Charge Density (EMCD) and Wave Form Propagation (WFP) Methods for Diagnostics of Engineering Materials, Components and Structures.

Inventors: Venu Gopal Madhav Annamdas and Pang Hock Lye John (School of MAE, NTU Singapore). Interview Date: 16 May 2013, Lab Demonstration Date: 22 May 2013. Status:  Patent Filed by A Star, [ETPL Ref.: SRC/Z/07571, NTU Ref.: TD/194/12.  Protection Period: 14 Months. A*STAR SERC Grant No.: 112 290 4014], Singapore. 

3. Research Scholarships & Grants from Various Schools & Colleges - Singapore: 

(a) Sch of CEE, NTU   2003-2006, (b) DSTA, Singapore Govt.  2006-08, (c) Sch of CEE, NTU 2014-2019 (d) Sch of MAE, NTU & SIM Tech, A Star    2012-2014  (e)Sch of EEE, NTU   2019-2020 (f) Energy Research Institute @ N, NTU 2024-2025

4. Research Scholar/Grant Dept of CEE, Univ of Pittsburgh 2008-09

2. The Silent Revolution in the Dashboard: Haptics through Piezoelectrics (Continental Automotive)

My journey took a fascinating turn, away from structural girders and subterranean tunnels, and directly into the sleek, high-tech world of automotive interiors. I joined Continental Automotive, a titan in the industry, with a mandate that felt both futuristic and profoundly practical: to revolutionize the human-machine interface in cars, specifically focusing on haptics applications for display screens.

The Problem: A Cold, Flat Future

The automotive industry was rapidly moving towards large, integrated display screens replacing traditional buttons and knobs. While visually stunning, these screens presented a critical problem: the loss of tactile feedback. Pressing a physical button provides an immediate, reassuring "click" – a confirmation that your input has been registered. A flat glass screen, however, offers nothing but a smooth, unresponsive surface. This lack of haptics in car displays was more than just an inconvenience; it was a safety hazard. Drivers had to take their eyes off the road to confirm interactions, increasing distraction. Furthermore, the premium feel and intuitive operation associated with high-end vehicles were being eroded by this "cold, flat future."

Continental faced the daunting challenge of bringing intuitive, safe, and premium tactile feedback back to the dashboard, without sacrificing the aesthetic and functional benefits of large digital displays. Traditional haptic motors were often bulky, power-hungry, and difficult to integrate seamlessly behind thin glass. We needed a solution that was precise, thin, energy-efficient, and could create a highly localized and customizable touch sensation directly on the display surface.

The Solution: Piezoelectrics - The Invisible Touch

This is where piezoelectric technology emerged as the silent, elegant answer. My role, spanning both the German and Singaporean arms of Continental Automotive, was to spearhead the development of cover lenses for these display screens that incorporated piezoelectric elements for haptic feedback.

Here's how we solved the problem:

1. The Piezoelectric Actuator as a "Micro-Motor": Instead of bulky mechanical motors, we leveraged the incredible precision and rapid response of piezoelectric ceramics. By integrating incredibly thin PZT elements directly into or beneath the display's cover glass, we could use them as miniature actuators.

2. Localized Vibration for "Feel": When an electrical signal was applied to these piezoelectric elements, they would rapidly deform, creating localized vibrations on the display surface. This micro-vibration, too subtle to be seen, was perfectly tuned to simulate the feeling of a click, a ripple, or even a texture under the user's finger.

3. Precise Control and Customization: The beauty of piezoelectrics is their ability to respond almost instantaneously and with extreme precision. This allowed us to:

   • Localize the haptic feedback: Only the specific area touched by the finger would vibrate, providing clear, unambiguous confirmation.

   • Vary the "feel": By changing the frequency, amplitude, and waveform of the electrical signal, we could engineer different tactile sensations – a soft bump, a firm click, a subtle buzz, or even a more complex texture. This allowed for rich, intuitive feedback that could be tailored to different functions or even personalized by the driver.

   • Maintain display aesthetics: The piezoelectric elements were so thin that they were virtually invisible, allowing for uncompromised display clarity and design.

   • Energy Efficiency: Compared to traditional haptic motors, piezoelectrics are remarkably energy-efficient for generating these micro-vibrations, critical for power-conscious automotive systems.

Crucially, my foundational knowledge in PZT, honed through years of academic research and lab work, proved invaluable in this corporate environment. The principles of electromechanical impedance, transducer interaction, and signal analysis were directly applicable. I wasn't just applying existing solutions; I was re-validating and adapting my deep understanding of piezoelectric behavior within the specific demands of automotive haptics. This intensive effort, a testament to the synergy between academic expertise and industrial innovation, took place within the Continental-NTU (Nanyang Technological University) joint corporate lab in Singapore. Our success in translating complex PZT physics into a practical, marketable solution for automotive applications culminated in a granted patent, solidifying our contribution to the next generation of in-car user experiences.

This wasn't just about adding a "buzz." It was about creating a seamless and intuitive human-machine interaction that brought back the essential tactile feedback lost with the rise of touchscreens. We moved from a cold, flat interface to one that was alive, responsive, and safer. My work at Continental involved not just theoretical understanding but the very practical application of piezoelectric principles to engineer this sensory revolution directly into the car's interior. It was a dramatic pivot for me, applying the fundamental science of piezoceramics to an entirely new, incredibly consumer-focused, and safety-critical application.

3. The Digital Frontier: My Leap from Smart Materials to AI-Driven Productivity

My career had been defined by the tangible world of smart materials and the precise physics of piezoelectric transducers. For years, I had focused on sensing the physical world – the integrity of concrete, the subtle vibrations of a car dashboard. But a new, even more profound challenge emerged, one that would redefine my professional trajectory and mark my most significant transition yet: the leap into Artificial Intelligence. This shift wasn't driven by a corporate mandate or an academic grant; it was sparked by a very personal inspiration.

My son, Shantanu Vasudev Krishna Annamdas, immersed in the digital realms of his video games, was constantly discussing the power of AI, how it governed game mechanics and character behavior. It was a revelation: if AI could manage the complex interactions of a virtual world, why couldn't it tame the overwhelming data flows of our real one? This insight, arriving just as my physical lab operations were winding down, struck me with such force that it prompted me to symbolically rename my lab – a declaration of my new direction.

The Problem: Email Overload – A Hidden Drain on Productivity

In the fast-paced automotive industry, just like many others, email had become a double-edged sword. While indispensable for communication, it had also transformed into a relentless torrent, a digital deluge that drowned productivity. Employees, particularly in complex, global organizations like those I'd worked with, spent countless hours sifting through inboxes, attempting to:

• Prioritize the critical from the trivial: Which emails demanded immediate attention, and which could wait?

• Categorize correctly: Was this a project update, a technical query, a supplier issue, or a legal concern? Manual sorting was time-consuming and prone to error.

• Manage the sheer volume: The constant influx led to cognitive overload, missed deadlines, and a pervasive sense of being "behind."

This was a hidden drain on efficiency, a silent enemy of productivity, and a significant environmental footprint due to unnecessary data storage and processing. Traditional email filters were rudimentary; they could flag spam, but they couldn't truly understand context, intent, or urgency. The automotive industry needed a sophisticated, intelligent solution to transform email from a burden into a powerful tool for enhanced productivity and even sustainability.

The Solution: AI-Driven Eco-Friendly Email Management

My deep-seated understanding of data analysis, pattern recognition, and system optimization, honed through years of working with complex sensor signals, proved to be unexpectedly transferable. I realized that the "signals" from an email – its sender, subject, content, attachments, and even its timing – were analogous to the sensor data I had meticulously analyzed for structural health. The solution lay in applying advanced AI techniques to bring order and intelligence to this digital chaos.

We developed an AI-driven Eco-Friendly Email Management system, built around two core components:

1. Email Multi-classification Model: This was the intelligence engine. Using advanced Natural Language Processing (NLP) and machine learning algorithms, the model was trained on vast datasets of automotive industry emails. It learned to automatically and accurately categorize incoming emails into multiple, predefined categories (e.g., "Critical Project Update," "Supplier Inquiry," "Technical Support," "Sales Lead," "Internal Admin"). This wasn't just keyword matching; it understood the intent and context of the email. This multi-classification approach meant an email could belong to several relevant categories, providing a richer understanding.

2. Email Ranking Model: Beyond classification, urgency was paramount. This model took the classified emails and ranked them based on a dynamic set of parameters, including sender importance, keywords indicating criticality, past interactions, and predefined user preferences. A "project deadline" email from a senior manager would instantly be ranked higher than a general newsletter.

The "Eco-Friendly" aspect was an integral part of the design. By precisely classifying and ranking emails, we dramatically reduced the need for employees to manually sort, search, and store unnecessary emails. This led to:

• Reduced server load and energy consumption: Fewer redundant emails being stored and processed globally.

• Optimized data management: Only truly relevant emails were retained for long periods, leading to a smaller digital footprint.

• Enhanced productivity: Employees gained back precious hours, focusing only on what truly mattered.

This project was a dramatic testament to the universal power of analytical thinking and AI. My transition from the physical domain of smart materials to the digital realm of AI was not merely a change of field; it was an expansion of my problem-solving capabilities, directly inspired by a child's interaction with virtual worlds, and ultimately leading to a solution with tangible benefits for efficiency and sustainability in the automotive industry.

4. From Smart Cities to Smart Capital: My AI Leap into Financial Forecasting

Just as the digital frontier of email management reshaped my perspective, another realm, equally complex and vital, beckoned: finance. The inspiration, once again, stemmed not from an academic paper or a corporate brief, but from the everyday brilliance of my son. His effortless navigation of predictive algorithms in gaming, his innate understanding of probabilities and outcomes, struck me with fresh force. If AI could predict player behavior, why couldn't it foresee the intricate dance of financial risk? This very personal spark ignited my foray into a critical challenge for the modern banking world: the prediction of loan default using AI modeling.

The Problem: The Unseen Costs of Trust

For any bank, the lifeblood of its operations is lending. But with every loan extended comes an inherent, often devastating, risk: default. Traditionally, banks rely on credit scores, historical data, and rigid criteria to assess risk. Yet, these methods often fall short. They are backward-looking, slow to adapt to new economic realities, and struggle to capture the myriad subtle indicators that truly signal an individual's or business's likelihood to default.

The consequences of unpredicted loan defaults are catastrophic. For a representative bank in Singapore, it means incurring colossal financial losses. These losses don't just impact profitability; they erode capital reserves, constrain lending capacity for legitimate borrowers, and can even destabilize the financial system. The challenge was immense: how could a bank move beyond reactive measures to proactively identify and mitigate the risk of default, transforming a blind leap of faith into an calculated, data-driven decision? The urgent need was for a system that could see the unseen, predict the unpredictable, and protect the financial stability of both the institution and its customers.

The Solution: AI as the Oracle of Creditworthiness

My answer was to bring the predictive power of AI, honed through years of analyzing complex sensor data and taming digital information flows, directly to the heart of financial risk management. I proposed and developed an AI-driven Loan Default Prediction model, designed to give banks an unprecedented foresight into their lending portfolios.

The core of our solution lies in a sophisticated AI model, accessible and transparently presented on GitHub: https://github.com/venuannamdas/Singapore_Bank_Loan_Project

This model doesn't just look at simple credit scores. It delves deep, analyzing a multitude of factors – financial history, behavioral patterns, macroeconomic indicators, and even subtle trends that human analysts might miss. Through advanced machine learning algorithms, it learns to identify complex, non-linear relationships that predict the probability of a borrower defaulting.

But prediction alone isn't enough; insights must be actionable. Therefore, we developed a powerful, interactive dashboard to visualize the impact of these predictions: 

raw.githack.com/venuannamdas/Singapore_Bank_Loan_Project/masterDataScience_AI_Banking_Project_Default_Details.html 

This dashboard is where the true power of the AI model comes to life. It vividly displays the financial loss incurred by a representative bank on a map, allowing stakeholders to immediately grasp the scale and distribution of potential risks. It empowers banks to:

• Proactively identify high-risk loans: Enabling early intervention, renegotiation, or strategic refusal.

• Optimize lending strategies: Directing capital to lower-risk, higher-return opportunities.

• Mitigate financial losses: By acting before defaults occur, saving vast sums that would otherwise be lost.

• Enhance regulatory compliance: Providing robust, data-driven rationale for lending decisions.

This project was a dramatic leap, applying the analytical rigor and predictive capabilities developed in my previous roles to an entirely new, high-stakes domain. It's about empowering financial institutions with the foresight to navigate risk, turning the tide from reactive loss mitigation to proactive financial intelligence – all inspired by the insightful world view of a young mind. 

5. Regression model to predict "median house value in California

 https://github.com/venuannamdas/mini_proj2/blob/master/2_1.ipynb

   Dashboard: House estimates at different locations.

   https://raw.githack.com/venuannamdas/12july/main/Housing%20Estimates%20in%20California.html

6. Stroke a social awareness page- (classified model)

https://github.com/venuannamdas/strokeXmeeti

Stroke info on the map, Vienna, Austria .

https://raw.githack.com/venuannamdas/strokeXmeeti/master/STROKE_VIENNA_MITALY_SINGAPORE_VENU.html

7. Covid19 info details of selected countries and Linear regression models

https://github.com/venuannamdas/mini_proj/blob/master/

On the map:

rawcdn.githack.com/venuannamdas/mini_proj/8a93dc51da3083e15e5ab6bf94de834c711c32dd/covid_data_map.html 

Guardians of the Urban Lifeline: Pioneering Real-Time SHM for Singapore's MRT

The very bedrock of modern civilization lies in its civil infrastructure. Yet, the safety and unwavering reliability of these colossal structures – the lifelines of our urban existence – remain paramount concerns, demanding constant vigilance. Nowhere is this more critical than for infrastructures involving immense human traffic and monumental investments, such as our beloved Mass Rapid Transit (MRT) stations, towering flyovers, and vital bridges. In a world where structural failures can lead to catastrophic consequences, the compelling need for high-quality, continuous online Structural Health Monitoring (SHM) is undeniable.

The Evolution of Smart Guardians: PZT Transducers at the Forefront

Over the past two decades, the quest for real-time, in-service, and truly "smart" material-based SHM methods has ignited a fervent interest among academic and industrial researchers globally. Amidst this intense innovation, the piezoceramic (PZT) transducer has emerged as a beacon of efficiency. This remarkable smart material, with its unique electromechanical coupling properties, has become the workhorse for two primary SHM techniques: the Electromechanical Impedance (EMI) method and Guided Ultrasonic Wave Propagation. My own research, deeply embedded in this revolution, has consistently pushed the boundaries of PZT applications.

The EMI models, in particular, have shown extraordinary promise in their ability to not only detect nascent damages within structures but also to discern and quantify the subtle or significant loadings acting upon them. This dual capability is crucial for a comprehensive understanding of a structure's health.

A Decade of PZT Prowess: DSTA's Vision and NTU's Execution

This book, a testament to years of dedicated effort, reviews some of the most significant advancements in the field of PZT-based SHM, charting the progress made over the past two decades. It specifically highlights the groundbreaking research carried out in load monitoring using the EMI technique, a vital project that benefited from crucial DSTA (Defense Science and Technology Agency) funding and was executed by Singapore's dedicated research team at Nanyang Technological University (NTU). This work directly fed into the critical needs of a nation acutely aware of infrastructure vulnerability.

A World First: Safeguarding Singapore's Subterranean Future

And then, the pinnacle – the moment where theoretical advancement met immediate, tangible impact. This book proudly presents what stands as the WORLD'S first application of PZT for MRT Excavation monitoring in Singapore.

This was not merely an academic exercise; it was a deployment in a high-stakes, real-world scenario where precision and reliability were non-negotiable. Excavation for MRT lines, especially in a densely populated and geologically complex urban environment like Singapore, carries inherent risks. The memory of past incidents underscored the necessity of unparalleled vigilance. By integrating PZT-based EMI monitoring into the excavation process, we were able to provide unprecedented insights into ground conditions and structural responses, safeguarding both the enormous investment and, most importantly, human lives. It was a testament to how decades of fundamental research in smart materials could culminate in a pioneering application that directly enhances public safety and reinforces the backbone of a nation's infrastructure.

This seminal work, encapsulating years of dedicated research and validated in the crucible of real-world application, can be explored further in:

• Journal Link: https://doi.org/10.1002/stc.418