Page 639 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

• Deploy agent-to-agent mesh synchronisation over local Wi-Fi.
• Formalise a shared knowledge-base schema (with ITU) for crowdsourced content.

3 Use Case Requirements 4.7: Education

• REQ-01: It is of added value that we have access to low-power single-board computer
(Raspberry Pi 4 or equivalent) with 512 GB SSD.
• REQ-02: It is of added value that we have access to solar panel + battery able to power
server for ≥12 h blackout.
• REQ-03: It is of added value that we have access to local 2.4 GHz/5 GHz Wi-Fi AP with
30 m indoor range.
• REQ-04: It is of added value that we have access to quantised NLP and recommender
models <200 MB.
• REQ-05: It is of added value that we have access to Arabic language support (tokeniser,
embeddings, TTS/STT).
• REQ-06: It is of added value that we have access to AES-256 device encryption and
local-only data storage.
• REQ-07: Each hub must run micro-agents exposed via REST/gRPC.
• REQ-08: Hubs must support LAN multicast or WebSocket for agent mesh sync.
• REQ-09: It is of added value that we have access to a knowledge-base stored as Markdown
+ vector embeddings (JSONL).

4 Sequence Diagram

5 References

[1] S. Y. Baroud and M. N. Qureshi, “Harnessing Industry 4.0: A Framework for Integrating AI
and Emerging Technologies,” in 2024 6th International Workshop on Artificial Intelligence and
Education (WAIE), 2024.

603