Itumo

Japan exhibitors
Real

Plant Show 2026
Innovation in Water Management
Booth number 2-F22

We have developed ABYSS, a proprietary AI model specialized in fluid analysis that combines physical laws with real fluid video data.

Using existing camera footage, ABYSS analyzes fluids such as water, oil, and chemical liquids to detect liquid levels, bubbles, foaming, scum, oil films, turbidity, leaks, and changes in flow conditions. This enables anomaly detection, early warning, and condition monitoring.

Unlike conventional approaches that rely only on statistical pattern matching, ABYSS is designed based on an understanding of fluid behavior, helping capture complex changes in real environments. It can also be introduced without installing additional sensors, since existing cameras can be used as “fluid sensors.” This helps reduce inspection labor, improve operational stability, support predictive maintenance, and enable unmanned monitoring.

At the exhibition, we will showcase four demo applications on monitors: foaming and scum monitoring for chemical plants, pipeline leak monitoring for the oil industry, flood risk detection through river monitoring, and liquid-state monitoring for pharmaceutical and biotech processes.

Exhibit Product

ABYSS
- Target Sector (English)
  Water treatment and wastewater facilities, chemical/material plants, oil and energy facilities, pharmaceutical/bio manufacturing, food, paper and textile plants, and other industrial sites handling water, oil, chemicals, slurry or cooling water. Designed for maintenance, environmental management, production engineering, utilities, quality control and DX teams that rely on visual inspection for foam, oil film, scum, turbidity, liquid level changes, leakage and flow-condition changes.
- Use Scenes (English)
  ABYSS monitors liquid surfaces in wastewater pits, separation tanks, discharge outlets, sedimentation tanks and flotation tanks through existing cameras, detecting foam, oil film, scum and turbidity. It can also be used to monitor flow conditions, liquid levels and viscosity changes in sight glasses and pipes, detect abnormal spray patterns or clogging in cleaning nozzles, and observe bubbles, dripping and liquid-condition changes in mixing, filling and coating processes. Outputs can be linked to dashboards, alerts, field notifications and APIs to reduce patrol inspections, missed events, and support predictive maintenance and process stability.
- Sales point (English)
  ABYSS is a fluid-specific AI solution that turns existing camera footage into data for monitoring liquid conditions. For water, oil, chemicals and other liquid processes, it captures qualitative state changes that are difficult to understand from conventional numeric sensors alone. Its proprietary physics-informed AI analyzes flow patterns, velocity vectors, viscosity, liquid surfaces, foam, oil film, scum and turbidity, supporting anomaly detection, predictive monitoring and first-level decision support. ABYSS can be introduced without installing large numbers of additional sensors, enabling non-contact, area-based monitoring that visualizes biased flow and stagnation. With continuous monitoring and alert integration, it helps reduce patrol work, environmental risk, unplanned equipment trouble and process instability.

Back