Inside aquAI - 21 Oct 2025 | Edition No. 14

Flow time: 5 min I your weekly pulse on AI news, tool and case studies reshaping the water sector

💡 Latest in AI: Google gives Gemini a map! 🌍 Now your AI knows where things actually are, perfect for smarter field ops, planning, and flood alerts.

🔬 AI Research Spotlight: Smarter water reuse 💧- AI is learning to turn wastewater into safe irrigation water while saving energy and costs.

🌊 Case Study: Spain’s SAIH system shows how real-time data and digital twins are reshaping how countries manage droughts and floods. 🇪🇸

🧠 Trending Tool: Recall is your AI “second brain” 🧩- summarize anything and build your own searchable knowledge hub.

🕵️ AI’s Shadow: MIT warns of the GenAI Divide ⚔️- billions spent, 95% no ROI, and workers secretly using AI that actually works.

Google launched Grounding with Google Maps in the Gemini API (Oct 17, 2025), letting developers connect Gemini’s reasoning to up-to-date Maps data from 250M+ places. Use it to return structured facts (addresses, hours, ratings) and even render an interactive Maps widget via a context token.

Grounding Gemini with Google Maps could be a game-changer for spatial intelligence in water management. Utilities, councils, and consultants can use real-time, map-linked AI to locate and visualize infrastructure, plan field operations, and communicate flood, drought, or contamination alerts with geographic accuracy. It enables AI agents to reason about catchments, treatment assets, and service zones using verified spatial data, reducing the risk of errors in reports, public dashboards, and planning tools. In short, it moves AI from abstract text models to location-aware decision systems that can better support resilience and response across the water network.

👉Full story

Researchers from the Harbin Institute of Technology and the University of Bologna reviewed how AI is transforming wastewater reuse and irrigation. They found that AI-driven monitoring, soft sensors, and smart aeration can cut energy use, improve effluent quality, and optimise irrigation scheduling using IoT weather and soil data. However, most models still focus on discharge compliance rather than reclaimed-water standards like the EU’s 2020/741. The study calls for more explainable, integrated AI systems that link wastewater treatment and agriculture, helping utilities and regulators build safe, efficient, and transparent water reuse networks.

This research shows that AI can close the loop between wastewater treatment and agriculture, unlocking safer and more efficient water reuse. By redesigning AI models to meet reclaimed-water standards and embedding transparency into decision systems, utilities can save energy, lower costs, and accelerate sustainable reuse programs that align with future water-security goals.

👉Full paper

Researchers from the University of Castilla-La Mancha and Polytechnic University of Valencia analysed how digitalisation, through IoT sensors, AI analytics, digital twins, and smart platforms is reshaping water management and policy in Spain. Their study examined impacts across the entire water cycle and at multiple management levels, highlighting clear benefits such as better monitoring, faster decision-making, and greater transparency, while also identifying barriers to wider adoption.

One example is SAIH (Sistema Automático de Información Hidrológica). Spain’s automated hydrological information system. SAIH operationalises real-time data for faster incident response, transparent water allocation, and coordinated flood and drought management across agencies. By integrating predictive tools and continuous monitoring, systems like SAIH go beyond optimising plant and network performance, they also strengthen regulation and governance through transparent, evidence-based decisions.

👉Full story

A “second brain” that summarizes YouTube, PDFs, articles, Google Docs, and podcasts, then auto-files everything into a self-organizing knowledge base with chat over your library, knowledge graph links, and spaced-repetition quizzes. Works via browser extension + web + mobile

🔌Try it

MIT’s Project NANDA reveals a striking truth: despite $30–40 billion invested in GenAI, 95% of companies see no financial return. Only a small fraction (≈5%) of pilots ever reach production. The real disruption isn’t happening through official channels, employees secretly use tools like ChatGPT or Claude to automate their daily work while corporate AI programs stall.

The study shows that the real innovation is happening outside IT governance. Employees are already crossing the GenAI divide using flexible, low-friction tools that adapt to their needs, exposing a widening gap between official enterprise systems (slow, rigid) and everyday productivity. This uncontrolled adoption brings both efficiency gains and serious data-security and compliance risks.

Organizations crossing the GenAI Divide succeed by adopting adaptive, learning AI systems that integrate deeply into workflows. This marks the rise of the Agentic Web, a new era where autonomous, memory-rich agents replace static tools, coordinating tasks across platforms. As companies lock in these systems, those that embrace learning, memory, and integration will lead the next wave of AI transformation.

👉Full story

Thanks for reading! I hope you’ve enjoyed this week’s edition and look forward to seeing you next week!

Dr. Andrea G.T