Singapore-based WaveScan Technologies, an A*STAR spin-off, specializes in the R&D of disruptive beamforming electromagnetics (EM)-based smart sensor systems and advanced AI algorithms. The semiconductor startup provides an end-to-end AI-enabled asset inspection solution, addressing the specific needs of the built environment sector.
As part of our newly launched semiconductor series, we spoke with WaveScan founder and CEO Dr Kush Agarwal.
Excerpts from the interview are below:
What sets WaveScan’s smart sensor systems apart from traditional non-destructive testing (NDT) solutions like X-ray, ultrasonic, or eddy current technologies?
WaveScan’s scanner systems propagate EM waves, similar to our mobile phones or Wi-Fi, which do not need physical contact with the surface for inspections (unlike ultrasonics) and work more accurately for non-metallic materials (unlike eddy current) due to their underlying science.
Also Read: Silicon Box’s Business Head on how chiplet architecture transforms semiconductor scalability
This enables us to capture high-resolution 3D images that are representative of structural information needed to identify the condition of the assets and hence make data-driven decisions to repair and maintain accordingly.
EM waves are non-ionizing and, therefore, completely safe when radiated under approved limits, making them easily deployable in buildings (residential and commercial) around people (unlike X-rays).
WaveScan’s technology enables high-resolution 4D sensing. Could you explain what 4D sensing entails and how it enhances NDT applications for built environments?
In 3D imaging technology, 3D usually corresponds to the spatial x, y, and z coordinates against the time axis. For WaveScan scanners, we capture the fourth dimension, i.e., velocity at any time. This fourth dimension becomes critical for specific applications in built environments, where we can identify loose tiles or fixtures of delaminating facades that are not visible to human eyes yet.
4D sensing can also help us further analyze specific images for structural inaccuracies that might go unseen otherwise. This is also useful for green energy assets where internal cracks and fractures can lead to unseen failures, such as wind turbine fan blades.
Beamforming EM and advanced AI are core to your system. How do these technologies work together to detect, analyze, and predict infrastructure defects?
Our scanning systems achieve their acquisition speed, spatial resolution, and advanced data collection capabilities using robotics-based automation by utilizing beamforming techniques. In the NDT industry, these parameters define the capabilities that enable specific use cases requiring sub-cm or sub-mm accuracy and turnaround time (i.e., the overall time of asset inspection).
Also Read: Zero-Error Systems: Safeguarding space travel from satellite collisions and debris
When scanners capture raw data, we use advanced radar signal post-processing techniques to generate high-resolution images, which are filtered and tagged using AI algorithms. For certain applications, we can also enhance 3D images using our AI tools, making capturing very small-scale defects possible.
With all these technologies working together, we perform 100 per cent scans of entire assets like a 20+ story building or a 100+ meters tall chimney stack, capturing every inch of its structural data—something that was not possible before using traditional NDT technologies.
Can you explain how WaveScan’s autonomous sensors and AI-enabled analytics solution create a predictive maintenance ecosystem for infrastructure?
Today, most inspections are reactive, meaning that either a professional engineer (PE) is conducting an assessment because of telltale signs of visible defects or an accident has occurred due to infrastructure failure.
The root cause of such reactive behavior in asset management is the lack of easy-to-use scanners, which makes inspections tedious and time-consuming. This has led to spot testing, which means random sampling across the entire asset to make a well-informed guess about its health.
With robotics-enabled automated scanners, WaveScan has reduced the setup and inspection times and made the holistic scanning of entire assets viable. This is crucial for detecting defects and potential failures in our built infrastructure. When millions of data points are collected, the data needs to be analyzed and tagged, which can be time-consuming and costly if done primarily by humans.
Hence, AI-enabled analysis helps to filter and analyze the data, classifying defects and tagging them under various categories. With automated data collection and AI-driven analytics, we’ve created a predictive maintenance ecosystem for infrastructure, among other industries.
What other sectors or industries could benefit from WaveScan’s advanced NDT solutions, and do you foresee any upcoming expansions?
Our scanning technology is already utilized in the Oil and Gas (O&G) industry, in addition to the built environment sector. Concrete-related infrastructure and corrosion mapping (through fireproofing materials) are some critical use cases where existing NDT technologies are not very reliable.
Our scanning technique has been deployed to minimize destructive testing and identify early-stage defects with high precision. It is also being used to develop other advanced NDT applications in green energy assets, such as inspections of wind turbine fan blades and carbon fiber tanks.
Skilled inspectors are typically needed for traditional NDT, especially for interpreting data. How does WaveScan’s AI-enabled defect analytics help reduce this dependence on skilled field inspectors?
In today’s inspection ecosystem, field inspectors are still needed to make the final judgment on the inspection data and reports. However, this is a very repetitive and tedious task that becomes more difficult as the inspection area increases.
With WaveScan’s AI-enabled image filtering, analysis, and tagging, the human analysis required by an expert is significantly reduced, and the NDT professional has to look mainly through the final tagged results and analysis reports. Hence, for a skilled inspector, our AI algorithms and automated report generation simplify their work, allowing them to focus on evaluating final reports and significantly save time, hence being cost-effective.
Finally, how do you see WaveScan’s work contributing to the broader goals of sustainable and predictive urban infrastructure management?
With our first-of-kind NDT scanners and a full suite of software solutions, we have been able to support our unique clientele, which consists of built environment stakeholders such as construction companies and asset owners, in scanning and diagnosing the holistic health of structures, restoring and upgrading landmarks, and conserving historical assets older than 100 years.
With a data-driven maintenance approach, the industry’s mindset is slowly evolving, and stakeholders want to make informed decisions about maintaining their assets. Our technology specs have also enabled building regulators to find solutions for NDT so that the industry can inspect and comply with various facade and structural inspection regulations. With technology, regulations, adoption, and awareness going hand-in-hand, the built environment industry would embrace our vision of predictive or preventive infrastructure management.
The post Transforming asset inspections: How WaveScan’s smart sensors and AI are shaping predictive maintenance appeared first on e27.