Eco-Bee: A Personalised Multi-Modal Agent for Advancing Student Climate Awareness and Sustainable Behaviour in Campus Ecosystems

2026-04-20 · Source: cs.AI updates on arXiv.org · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Environmental Science & Earth Systems · Depth: Intermediate, medium

Summary

Eco-Bee is a multi-modal AI agent designed to enhance student climate awareness and promote sustainable behavior within university campuses. It integrates large language models, specifically Mistral's Pixtral-12B, with the Planetary Boundaries framework, translated into an "Eco-Score." The platform provides personalized, actionable insights, peer benchmarking, and gamified challenges to students, connecting their daily choices to broader environmental limits. A pilot study involving 52 students across multiple campus networks demonstrated strong positive reception, with 96% supporting a campus-wide rollout and 80% willing to engage with peer benchmarking. Students reported a clearer understanding of how their behaviors impact planetary limits, with average Eco-Scores around 50.9 out of 100.

Key takeaway

For university administrators and sustainability program managers seeking to drive measurable climate action among students, Eco-Bee offers a scalable, AI-mediated solution. You should consider implementing a similar multi-modal agent that provides personalized, boundary-aligned feedback and gamified challenges. This approach can significantly increase student engagement and understanding of their environmental impact, fostering sustained behavioral change and supporting broader campus sustainability goals.

Key insights

Eco-Bee uses AI and planetary boundaries to personalize sustainability feedback and drive student behavioral change.

Principles

• Personalized, multi-boundary feedback is more effective than generic carbon metrics.
• Gamified and socially reinforced engagement fosters motivation.
• Privacy by design is crucial for user adoption.

Method

Eco-Bee employs a multi-modal intake, Vision AI (Pixtral-12B) for image classification, a scoring engine mapping behaviors to nine planetary boundaries, a node2vec-based recommendation engine, and a chatbot service, all deployed on Vercel with a Next.js frontend and Python FastAPI backend.

In practice

• Integrate multi-modal AI for comprehensive behavior analysis.
• Utilize the Planetary Boundaries framework for holistic impact assessment.
• Implement gamification and peer benchmarking for sustained engagement.

Topics

• Eco-Bee
• Planetary Boundaries Framework
• Multi-modal AI Agent
• Sustainable Behavior Change
• Gamification & Peer Benchmarking

Best for: AI Scientist, AI Engineer, Research Scientist

Related on AIssential

• LectūraAgents: A Multi-Agent Framework for Adaptive Personalized AI-Assisted Learning and Embodied Teaching — LectūraAgents provides adaptive, personalized, embodied AI teaching via a hierarchical multi-agent framework.
• Agent Swarms and Knowledge Graphs for Autonomous Software Development with Siddhant Pardeshi - #763 — Autonomous development systems can deliver production-ready software at enterprise scale by orchestrating AI agent swarms.
• AIhub monthly digest: February 2026 – collective decision making, multi-modal learning, and governing the rise of interactive AI — AIhub's February 2026 digest highlights progress in multi-agent systems, robot learning, neurosymbolic AI, and governance.
• A Multi-Objective Optimization Approach for Sustainable AI-Driven Entrepreneurship in Resilient Economies — The EcoAI-Resilience framework optimizes AI deployment for sustainability, economic resilience, and environmental cost reduction.
• ClimateAgents: A Multi-Agent Research Assistant for Social-Climate Dynamics Analysis — Multi-agent AI systems can augment interdisciplinary research by coordinating specialized agents for complex socio-environmental analysis.
• Nectar Social: Rewriting the Modern Marketing Playbook — AI agents are essential for brands to maintain presence and influence buying conversations in private social channels.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.AI updates on arXiv.org.