Skip to content
Ernest Obimpeh

Laboratory scientist and developer, Accra, Ghana

Same mistake.
Same problem?

I studied laboratory science. Somewhere along the way, I fell in love with computing. These days, I am interested in how computers can help us learn, do science and make sense of information.

Student ARinseFill buretteAdd indicatorTitrateReadRecordStudent BRinseFill buretteAdd indicatorTitrateRetryRetryRecord

Student A. Smooth procedure, wrong value recorded. A conceptual gap?

Student B. Repeated adjustments before the same error. Trouble with the instrument?

Illustrative concept, not real student data.

A question I keep coming back to

Could the sequence of actions leading to an experimental error, read alongside an independent assessment of conceptual understanding, help us distinguish between different learning difficulties? And would visualizing that sequence help a teacher choose the right kind of help?

Imagine two students getting the same reading wrong in a titration. One understands the chemistry but struggles with the burette. The other can operate the burette but does not understand the chemistry. Both may lose the same mark. How do we know which student needs what kind of help?

Elementa already records timestamped procedural errors, attempt histories and skill scores, and compares students’ recorded titration readings against actual instrument values in assigned practicals. Since November 2026 it also records each student’s actions at the bench in order, and whether each judged step was right the first time.

The titration bench can already tell a misread burette from a titration stopped at the wrong point. What it cannot yet tell is why a student stopped where they did. A fuller research version, with consent and a separate concept test, is designed and will not run until it has ethics approval.

Education is where much of my current work lies, but it is not where my curiosity ends. I would love to work with scientists in different fields and investigate how AI and interactive tools could help them do their work better.

From ideas to actual projects

2025 to present

Founder and lead developer

Elementa Labs Africa

I still remember sneaking out of school in the evenings to attend chemistry practicals at another school because we had very little access to a laboratory at my own school. Years later, I started building a virtual lab for students who face a similar problem.

Early rollout, as of November 2026

Students
81
Teachers
15
Acid-base titration
346
Microscope familiarization
326
Onion cell microscopy
228
Simple pendulum
104
Stomata
72
Sessions recorded in the five most-used practicals. These counts include some teacher demonstrations and say nothing yet about learning.

Visit elementalabsafrica.com

2026

Founder and developer

PINN Ghana

What if we had a better way to collect and make sense of information about our roads in Ghana? PINN is my attempt to explore that question.

Visit pinnghana.com

2018

Undergraduate dissertation

Pesticide residues in dried chilli pepper

For my final-year project, I travelled to Brewaniase, Kwadaso, Sunyani and Winneba to collect dried chilli pepper from farmers. I wanted to find out whether the peppers contained pesticide residues. The Ghana Standards Authority provided technical assistance.

Samples
120
Pesticides
36
Regions
4

How I got here

University of Cape Coast

BSc Laboratory Technology

2014 to 2018

Around 2016, while studying laboratory technology, I became interested in writing software. I did not even have a laptop at the time, but I had plenty of ideas. One was a way to identify plants from a photograph; another was a way to find misplaced keys. I later discovered that people were already working on such things. That made me even more curious about computing.

Building Elementa has made me appreciate how much there is to learn beyond getting software to work. I now want to study these questions more rigorously through doctoral research.

Let’s talk science and technology

Human-computer interaction, artificial intelligence, interactive data visualization, learning analytics and computing for scientific discovery.

These are the things I think about most. I enjoy talking about them, especially with people who know things I do not. If you are working on something related, I would be happy to hear from you.

Email meCV (PDF)GitHub