Portable screen-printed carbon electrode (SPCE) biosensors provide a fast and cost-effective solution for detecting microcystin-LR (MC-LR), one of the most toxic compounds released by cyanobacteria during harmful algal blooms in freshwater ecosystems. Even at very low concentrations, MC-LR poses serious health risks, including liver damage, and has been associated with a higher likelihood of liver and colorectal cancers. To protect public health, the World Health Organization recommends that drinking water contain no more than 1 microgram of MC-LR per liter.

These biosensors estimate toxin levels by monitoring changes in electrochemical signals generated when MC-LR is present. However, their performance is often influenced by the characteristics of the water sample itself. Variations in pH, turbidity, electrical conductivity, and other water quality parameters can alter sensor responses, making it necessary to recalibrate the device for each individual sample.

A research team from Hanbat National University in South Korea and the University of Central Florida in the United States has developed a machine learning-based approach that overcomes this limitation by compensating for differences in water quality. The study, led by Professor Jungsu Park of Hanbat National University and Professor Woo Hyoung Lee of the University of Central Florida, was first published online on 26 March 2026 before appearing in Volume 298 of Water Research on 15 June 2026.

According to Professor Park, "This work provides a robust data-driven framework for characterizing biosensor-water matrix interactions and offers a practical approach to improving the speed and accuracy of on-site MC-LR detection in complex environmental waters."

To develop the predictive model, the researchers gathered 201 datasets from 27 sampling locations across Florida, covering freshwater, estuarine, and transitional water bodies with diverse environmental conditions. For every sample, they recorded key water quality indicators—including pH, turbidity, electrical conductivity, total dissolved solids, ultraviolet absorbance at 254 nanometers (UV254), and the biosensor's electrochemical impedance (Z'), which varies with MC-LR concentration. These measurements served as model inputs, while laboratory-determined MC-LR concentrations were used as the target outputs.

The team evaluated several machine learning algorithms, with the Extreme Gradient Boosting (XGBoost) model delivering the strongest performance. It achieved a Nash-Sutcliffe efficiency of 0.89 and a root mean square error of 13.21, demonstrating that a single generalized model could accurately estimate MC-LR concentrations across a wide range of water conditions without the need for sample-specific calibration.

To better understand how different variables contributed to the model's predictions, the researchers employed Shapley Additive Explanations (SHAP), an explainable artificial intelligence technique. Their analysis revealed that the biosensor's electrochemical impedance had the greatest influence on prediction accuracy, followed by electrical conductivity, pH, UV254 absorbance, and turbidity. These findings highlight the importance of incorporating water quality characteristics into the predictive framework to improve biosensor reliability.

Professor Park noted, "This framework eliminates the need for repeated sample-specific calibration, reducing time, labor, and sensor consumption. Compared to conventional workflows, it can reduce sensor usage and thereby lowering cost and environmental burden while improving analytical efficiency."

With climate change contributing to an increase in the frequency and severity of harmful algal blooms, this machine learning-assisted biosensing approach has the potential to make monitoring of toxic cyanobacterial contaminants faster, more accurate, and more practical for routine testing of drinking water supplies and recreational water bodies.

Telecom providers are pushing for fast 6G standardization and quick adoption across enterprise and consumer markets, with AI playing a central role.

The transition to 6G marks a major shift — potentially becoming the first generation of wireless networks built to be AI-native. Because AI will deeply influence how 6G operates, engineers face a new challenge: validating systems that are far more adaptive, intelligent, and fast than previous generations.

This overview outlines how AI can support 6G design validation for teams in communication service providers, mobile operators, technology vendors, and device manufacturers. It also explores emerging applications enabled by AI and 6G, the types of AI techniques involved, and how these tools can streamline design and testing workflows.
 
What new opportunities will 6G and AI unlock?
AI and 6G together are expected to drive major innovations, including real-time digital twins, advanced manufacturing systems, highly autonomous transport, holographic communication, and widespread edge intelligence. These capabilities align with the visions of the ITU and 3GPP for 2030 and beyond.

Real-time digital twins
With widespread coverage, extremely low latency, and high throughput, 6G paired with AI could create high-fidelity, real-time digital counterparts of physical assets and environments. These digital twins would support modeling, control, analysis, and simulation with unprecedented accuracy. Digital twin networks could mirror actual network conditions to enable continuous optimization, especially when combined with integrated sensing and communication (ISAC).

Smart factories
AI-enhanced 6G connectivity could enable industrial automation at scale through reliable, ultra-responsive data exchange across robotics, industrial IoT, and intelligent devices. “Industrial 6G” may enable fully automated operations in environments such as factories, ports, and airports, supported by private 6G deployments.

Autonomous mobility
Next-generation mobility systems — from autonomous vehicles to intelligent transportation — will rely on AI-powered 6G capabilities. This includes AI-assisted driving, real-time mapping, and precise positioning for cellular vehicle-to-everything (C-V2X) interactions.

Holographic communication
Future 6G and AI infrastructure may support immersive communications such as holographic telepresence and multi-sensory remote interaction. AI-driven semantic communication could reduce bandwidth demands by transmitting only the essential meaning behind data-heavy content.

Distributed edge intelligence
6G is expected to blur the line between communication and computing by pushing AI models to the network edge. This could enable coordinated inference, collaborative robotics, and pervasive, real-time intelligence across devices.
 
How will AI improve 6G network design and operations?
6G will involve both physical elements (e.g., radios, base stations, user devices) and logical components (e.g., RAN, core network functions, protocol stacks). Many of these will be optimized using AI during design, validation, and even runtime.

AI-native air interface
AI could enhance key radio functions such as channel estimation, symbol detection, beam selection, modulation, and antenna configuration. These models may operate on devices, at the base station, or jointly across both.

AI-assisted beamforming
AI methods may support:

AI-optimized RAN
AI could enable a self-organizing RAN capable of real-time adaptation, end-to-end optimization, and autonomous performance tuning.

Automated network management
AI-driven operations may include predictive maintenance, traffic forecasting, energy optimization, and intelligent resource allocation. Real-time threat detection and mitigation could also be enhanced through AI analytics.
 
Which AI techniques are most useful for validating 6G performance?
A range of AI methods — deep learning, reinforcement learning, generative models, and more — will support system-level design and testing.

Reinforcement learning (RL)
RL is well-suited for automating decision-making in unpredictable environments and may be applied to:

Deep neural networks (DNNs)
DNNs may support tasks such as:

Transformer models
Transformer autoencoders may enhance CSI compression and feedback efficiency.

Graph neural networks (GNNs)
GNNs can model network topology and spatial relationships for interference control, mobility forecasting, and resource allocation.

Generative adversarial networks (GANs)
GANs can generate realistic channel data, support denoising, and detect anomalies.

Large reasoning/action models
These emerging agentic models may coordinate complex workflows and help test sophisticated, multi-component 6G systems.
 
How will synthetic AI data support 6G testing and validation?
AI-generated data will be crucial for exploring the huge range of possible 6G conditions — many of which cannot be physically tested early on.

Key synthetic-data methods include:

 
Can AI help validate 6G hardware and chip designs?
Yes. AI-powered anomaly detection, automation, and data-driven modeling could support the design of components for sub-THz frequencies, UM-MIMO, and other 6G features.
Key methods include:

 
What challenges come with using AI for 6G validation?
AI’s reliability isn’t guaranteed. Issues include out-of-distribution errors, limited data, low interpretability, overfitting, and hallucinations. To improve trustworthiness:

 
Keysight’s role
This summary illustrates how AI can support 6G design and testing. Keysight provides tools, research expertise, and 6G-ready test solutions to help engineering teams innovate with confidence throughout development.

Click here to know more.