The company demonstrated how multispectral sensing, on-device AI and cross-system analysis can help data centre operators identify emerging electrical, thermal and cooling risks before they develop into operational incidents.

SINGAPORE, Sept. 30, 2026 /PRNewswire/ -- Shenzhen HQVT Technology Co., Ltd. ("HQVT" or the "Company", stock code: 1392.HK) showcased its multispectral AI solutions for data centre safety, including the HQ 6000 Multispectral AI Terminal and the In-Cabinet Multispectral AI Terminal, at Data Centre World Asia 2026, held from 29 to 30 September at Marina Bay Sands in Singapore.

Also on display was HQVT's Privacy-Preserving Multispectral AI Terminal. Designed for privacy-sensitive environments, the terminal uses only ultraviolet and infrared sensing, without visible-light imaging, enabling continuous anomaly detection without capturing conventional video imagery.

Through its data centre portfolio, HQVT demonstrated how multispectral sensing can complement existing Data Centre Infrastructure Management (DCIM), helping operators identify early-stage physical anomalies that conventional monitoring systems may miss.

Seeing Risk Before It Escalates

As artificial intelligence workloads drive higher computing density, power consumption at individual cabinets is rising sharply. This puts greater pressure on power distribution, cooling and day-to-day operations. Small physical changes—including abnormal temperature rise, electrical discharge and uneven heat distribution—may begin inside enclosed cabinets or develop intermittently under changing loads, making them difficult to identify through periodic manual inspection or conventional visible-light cameras alone.

This raises a practical question for data centre operators: how can a facility detect the earliest physical signs of risk, rather than waiting for a threshold alarm or equipment failure?

HQVT approaches the problem by combining sensing across ultraviolet, infrared and visible-light spectrum bands with on-device AI analysis. Instead of relying on a single image or isolated alarm, the system is designed to continuously examine changes in equipment condition and provide additional evidence for risk assessment.

At Data Centre World Asia, HQVT presented the HQ 6000 and its In-Cabinet Multispectral AI Terminal for data centre applications. The HQ 6000 Multispectral AI Terminal forms part of HQVT's general-purpose multispectral sensing portfolio for data centre environments. The In-Cabinet Multispectral AI Terminal brings continuous detection closer to critical components inside electrical and server cabinets, where early signs of abnormal heat or electrical discharge can otherwise remain hidden.

Together, the configurations allow multispectral AI sensing to be deployed according to the physical layout, equipment density and operating requirements of different data centre zones.

From Isolated Alerts to Facility-Wide Insight

Detecting an abnormal signal is only the first step. Data centre operators also need to understand whether the signal represents a persistent risk, how quickly it is developing and which asset requires attention first.

HQVT's data centre solution is built around a Three-tier Collaborative Multispectral AI Perception System:

• In-cabinet sensing continuously senses local abnormal signals at critical locations inside cabinets.
• Spatial sensing senses abnormal changes across data-centre aisles and other critical areas.
• Mobile inspection extends sensing to routes and locations that are difficult to cover with fixed devices.

Data collected from these sensing points can be analysed alongside equipment status, historical trends and other facility information through HQVT's Zhiyuan Origin Large Model. The platform is designed to support complex-event correlation, confidence scoring and risk grading, helping operators move from fragmented alerts towards a more complete assessment of facility conditions.

Rather than replacing existing DCIM platforms, the solution is intended to add a multispectral physical-sensing layer. This gives existing management systems access to additional information on power-distribution anomalies, cabinet thermal conditions, cooling-system irregularities and other operational risks.

"Data centre safety increasingly depends on whether operators can identify small physical changes before they become major incidents," said Roger Tian, sales director at HQVT. "Our aim is not simply to add another stream of alarms. We want to help operators connect multispectral signals with equipment conditions and operational context, so that they can understand where a risk is emerging, whether it is developing and what should be addressed first."

Supporting Resilient Data Centre Growth in Southeast Asia

During the two-day event, HQVT engaged with data centre operators, engineering teams, technology partners and other industry stakeholders from Singapore and across Southeast Asia. Discussions focused on the practical deployment of multispectral sensing in high-density computing environments, as well as its integration with existing facility-management platforms.

HQVT's booth attracted a steady flow of visitors throughout the event, with many stopping for detailed discussions on how multispectral AI can reveal physical risks that may remain invisible to conventional monitoring systems. Visitors showed particular interest in the terminals' ability to identify early signs of abnormal heat and electrical discharge, as well as the potential to integrate these capabilities into existing data centre operations. The exchanges reflected growing demand for more continuous and multidimensional approaches to facility safety as computing density continues to rise.

For fast-growing data centre markets in Southeast Asia, the challenge is not only to build more capacity, but also to operate that capacity reliably under demanding power, cooling and environmental conditions. HQVT believes that stronger physical sensing can support this transition by helping facility teams discover early-stage risks, reduce inspection blind spots and make maintenance decisions with more complete evidence.

Following Data Centre World Asia 2026, HQVT will continue to engage with regional customers and partners to explore local deployment, system integration and scenario-based applications of multispectral AI across data centres and other critical infrastructure environments.

About HQVT

Founded in 2013 and headquartered in Shenzhen, China, Shenzhen HQVT Technology Co., Ltd. is a multispectral AI technology company listed on the Main Board of The Stock Exchange of Hong Kong Limited under stock code 1392.HK.

Based on its "Optics-Sensor-Imaging-Computing" technical architecture, HQVT has developed multispectral perception capabilities spanning ultraviolet, infrared and visible-light spectrum bands. Its products and solutions are used in data centres, power systems, energy facilities and other high-value applications. To date, the Company has served more than 2,500 customers.

According to Frost & Sullivan, HQVT ranked first by revenue among multispectral AI companies in China and among multispectral AI large model service providers in China in 2025.

For more information, please visit: https://www.hqvt.com/en/