April - June Newsletter

The recent months have been eventful and nothing short of captivating. We've gathered some of our highlights into this edition, and we couldn't wait to share them with you. Have a read!

OpenSense 2026

Samuel Mbugua, a researcher at our centre, had the incredible opportunity to present our work on Automating Quality Control of rainfall telemetry via the fusion of CMLs and hourly multivariate ERA5 data at the second International Conference on Opportunistic Sensing of Precipitation (OpenSense 2026).

Hosted at the Royal Netherlands Meteorological Institute (KNMI) in De Bilt, the Netherlands, the conference brought together global experts dedicated to advancing "opportunistic sensing" — using unconventional tools like commercial microwave links (CMLs) and personal weather stations to provide reliable, high-resolution precipitation observations where traditional networks fall short.

Across Sub-Saharan Africa, weather monitoring networks are extremely sparse, creating what we call the "Clog vs. Correlation Paradox". During highly localized Sahelian storms, it is mathematically impossible for traditional quality control systems to distinguish between natural weather variance and actual mechanical failures (like ATMOS 41 rain gauges clogged by dust and debris). To solve this, we introduced a zero-hardware "Virtual Sensor". By fusing opportunistic cell tower signals (CMLs) with ERA5 atmospheric thermodynamics, we built a physics-informed Virtual Sensor using a Luenberger Observer.

This work underpins the essence of having scalable frameworks that empower African meteorological agencies to bypass the limits of sparse networks, confidently monitor station health, and protect the high-fidelity historical weather records essential for long-term climate resilience.

A huge thanks to Austin Kaburia and Ciira Maina for co-authoring, and our partners too for being part of this: TAHMO, KENET, International Development Research Centre (IDRC), and Foreign, Commonwealth and Development Office.

DeKUT Staff Sports and Fun Day

Work with no play, makes Jack a dull boy. Away from our work posts, we took part in the staff sports and fun day at the Dedan Kimathi University of Technology. Through the day’s activities we interacted with colleagues from other departments and didn't shy off exemplifying great sportsmanship. Yuri Njathi clinched victory in chess, Daisy Munga cycled her way into second place in the bicycle race and we also won in football having been part of the ICT team. The day was a success and we extend our gratitude to the DeKUT fraternity for pulling this off.

Viva DeKUT!

Publication

RUL Prediction in LFP Batteries: Comparison of Gompertz, LSTM and Gompertz-Informed LSTM Models for Interpretability and Accuracy

Batteries are a reliable source for alternative energy and for long-term adoption in electric vehicles and large-scale energy storage predicting battery lifespan boosts reliability.

In our recent publication by Yuri Njathi, Ciira wa Maina and Edwell T. Mharakura, we focus on predicting the Remaining Useful Life (RUL) in LIthium Iron Phosphate batteries which dominate the Battery Energy Storage System (BESS) market. We focus on the integration of the Gompertz function into deep learning architectures. Even when pure data-driven models are more accurate, a physics-informed approach provides an essential confidence monitor. By identifying the stabilization point of battery parameters at 85% SoH, BESS operators can use Gompertz-based mathematical signals to transition from provisional estimates to high-confidence RUL predictions.

Read the publication here.

Minds, Machines and Matrices (M3) Workshop

From 6th to 10th July, we shall be hosting the Minds, Machines and Matrices workshop at the Dedan Kimathi University of Technology. We aim at equipping participants with data science and machine learning fundamentals through hands-on sessions leveraging real-world applications. For our online audience, we have made room for you. Further communication on the same will be made on our socials.

Climate Informatics Conference 2026

Prof. Ciira Maina delivered a keynote talk at Climate Informatics 2026 in Lausanne, Switzerland. The talk titled “Forests, birds and weather: Using AI to make sense of our changing world” described our unified approach to environmental and climate research from bioacoustics, to forest monitoring and weather forecasting.

Blogs

In our recent blog, “The forest is speaking, are we finally listening?”, Braxton Mandara elaborates on why the fight against silence in our ecosystems is crucial, and the need of developing low-cost ecosystem monitoring tools in combating this. The future of conservation will not be determined solely by how many trees we plant or how many hectares we restore. It will also be determined by our ability to understand ecosystems before they reach a point of irreversible decline and depends on whether we can build monitoring systems that reflect local realities, empower communities, and reveal change while there is still time to act.

Tutorials

For data science and machine learning enthusiasts, mastering the fundamentals is a continuous journey. We have made learning material accessible on our website through practical tutorials, each featuring a corresponding notebook allowing you to master the concepts through hands-on examples. Whether you are starting out or are already conversant with machine learning, the tutorials will advance your expertise. Explore our tutorials here.

Graduation

The 20th graduation ceremony at the Dedan Kimathi University of Technology ushered our very own James Muiru and Zipporah Kavutha into the next big adventure. Their dedication and perseverance in excelling academically has surely paid off. Having completed their undergraduate studies, they are ready to grace the professional work environment. Join us in celebrating them on achieving this milestone. Congratulations! You have an incredible future ahead.

Visits

We had the honour of hosting a number guests including:

•

Martin Muriuki and Gabriel Kiarie from Natural State [linkedin post]

•

Prof. Damien Hanyurwimfura, Prof. Kizito Nkurikiyeyezu, Myriam Mujawamariya, and Richard Musabe from TAIRI lab at University of Rwanda [linkedin post]

•

A team from the Distributed AI Research Institute (DAIR) led by Timnit Gebru [linkedin post]

•

Prof Vukosi Marivate from African Institute for Data Science and Artificial Intelligence (AfriDSAI) [linkedin post]

•

Loise Ochanda, Dr. Emmy Chirchir and Bomi Okuyiga from IDRC [linkedin post]

Through our engagements we shared the work that we do, exchange ideas and have a good belly laugh.

Dedan Kimathi University of Technology - Nyeri, Kenya

Subscribe to our newsletter here to stay updated on what we do.