Page 45 - AI for Good-Innovate for Impact

AI for Good-Innovate for Impact

4. Reduce lead time for innovations from research labs such as WINEST which in turn creates
a strong research mentality, which is essential for industrial and wireless infrastructure.

Using our proposal, contributors from emerging regions such as Africa are able to create
impactful and collaborative contributions to ITU and other standards bodies. 6 - FUT

Partner name: Nigerian Communications Commission
Background papers and/or references: Source Code

6�2�2 Future work

Data collection, Proof of concept development, Model development, Setup reference tools,
notebooks and simulation environment Elaborate proposal: Building upon our extensive work
within the ITU ML5G initiative and the insightful presentations made during the ITU workshop in
Geneva in July 2023, we, the WINEST team, led by Prof. Agajo, propose an innovative project
that uses AI to foster collaboration and reduce 6G standards barrier for African contributors.
If we are given the scholarships and resources:

1. We would continue our effort in the open source dataset in ITU Build-a-thon – we have
been actively developing on this dataset as presented in the most recent ITU FGAN Build-
a-Thon workshop on 19 Jan 2024
2. We would continue our efforts in annotating and creating the HF dataset and reference
tools are available from dataset link and source code link.
3. We would also continue our collaboration with NCC (Nigerian Communications
Commission) and other regional standards bodies such as ITU SG13 RG AFR.
WINEST team would continue with the above effort led by Prof. Agajo who has already made
international presentations in July 2023 in Geneva during ITU workshop. Our students such
Ms. Yemisi and Mr. Blessed have already been featured in ITU perspectives videos which can
be found here.

6�3� Use case requirements

This section describes the requirements for each entity on using AI to reduce the 6G standard
barrier for African contributors use case.
• REQ-01: It is required to perform NLP parsing on the Raw data
• REQ-02: It is required to annotate the data as a step to preparing the data for fine turning/
training
• REQ-03: It is recommended to use validated responses for the fine turning
• REQ-04: The Potential innovator require the inferred knowledge to generate responses
on 6G innovation

29