Skip to content

Benedict Quartey ’18: A Small Autonomous Vehicle, a Larger Field of Possibility

Benedict Quartey is in front of a white background. He is smiling and wearing a pale pink shirt.

From an affordable research vehicle to Ghanaian-language technology and robots that can learn, Benedict Quartey’s work connects scientific questions with the opportunity to pursue them.

In March 2019, a design workshop in Saint-Étienne, France, was built around a small autonomous vehicle developed at Ashesi. Participants were invited to use RollE to explore how a moving vehicle could gather information from its surroundings. The workshop brought a research platform conceived in Ghana into a different community of designers and researchers. It also gave its creator, Benedict Quartey ’18, another setting in which to share the work, including a keynote at the city’s International Design Biennale on making autonomous-vehicle research more accessible.

RollE had begun with a practical obstacle. Researchers interested in autonomous vehicles needed equipment on which to test their ideas, but cost could put that equipment beyond reach. Working with Dr. Ayorkor Korsah, Quartey developed an affordable, modular platform that could collect driving data under remote control and use a neural network to drive autonomously. Their co-authored paper described a tool intended to make experimentation more accessible to students and researchers, including those working with limited resources in Africa.

Two young men in a tiled corridor watch a small robotic car sitting in a lane marked with white tape.
RollE follows a taped test lane in an Ashesi corridor in 2018.

For Quartey, the project also opened an unexpected career. He had completed a banking internship in London and was considering continuing in that field. “When I was at Ashesi, I wasn’t thinking about a PhD,” he recalled. His capstone introduced a different possibility: the intellectual work of pursuing a question, testing an approach, and contributing something others could examine and build on.

Dr. Korsah encouraged him to document the project carefully and share it at conferences. The collaboration produced a research paper, while an assistantship at Ashesi gave him time to continue investigating beyond graduation. That period helped turn an engaging undergraduate project into a path toward doctoral research. The invitation to Saint-Étienne was one indication that the work could have a life beyond its original setting.

His next work brought the question of access closer to everyday life. At Nokwary Technologies, Quartey worked on speech-driven financial services for African languages. Here, the challenge was not the price of research equipment but whether people could use a digital service in a language they understood. His work as a founding research-and-development engineer included technology for voice-based banking and a deployment partnership with Ecobank.

That effort connected with research underway at Ashesi. A team led by Dennis Asamoah Owusu ’12 and David Sampah, with collaborators including Dr. Korsah and Quartey, developed language resources for Ghanaian-language technology. In 2022, researchers at Ashesi and Nokwary released speech datasets in Asante Twi, Akuapem Twi, Fante, and Ga, to their knowledge the first openly available speech data for any Ghanaian language. The purpose extended beyond one product: other researchers and developers could use the data to investigate speech technology and build applications of their own.

The datasets also opened scientific questions. How much training data does a speech system need to be useful in practice? Can recordings focused on a particular activity, such as financial transactions, support broader applications? The team set out these questions when it released the data. A service designed around an immediate need was also helping create a shared resource for further inquiry.

At Brown University, Quartey’s doctoral research brought language and physical action together in another way. Working with George Konidaris and collaborators, he investigated how robots could interpret instructions, understand their surroundings, and complete unfamiliar combinations of tasks. The underlying difficulty is easy to recognize: knowing how to perform several actions does not necessarily mean knowing how to combine them when the goal changes.

Benedict Quartey speaks at a podium between two large screens showing his title slide, with an audience at round tables.
Quartey presents research on robots that follow complex instructions at the 2024 Northeast Robotics Colloquium at UMass Amherst. Photo: courtesy of Benedict Quartey

In a paper accepted to IROS 2026, a major international robotics conference, Quartey and his collaborators developed skills that predict both an action and the changes it should produce. Those predictions help a robot plan combinations of skills it has already learned, rather than being retrained each time the combination changes. It is a step toward machines that can make better use of experience when they encounter something new.

That line of inquiry now informs a further ambition. Quartey has co-founded Primordiant, previously called SuperRadiant, which joined Y Combinator’s Fall 2026 batch. The company is developing robots intended to connect scientific reasoning with physical experimentation: observing a situation, proposing an explanation, and testing it. Its ambition moves the robot from carrying out a prescribed task toward participating in the process of discovery.

Three men stand with arms folded behind a Y Combinator sign outside a white building.
Quartey (center) with his Primordiant co-founders, Owen Tower and Cooper Niu, at Y Combinator. Photo: Y Combinator

The route has not been simply outward. In 2023, Quartey gave a lecture in Ashesi’s robotics course on how robots understand and follow instructions. He chaired faculty engagement for the 2024 Deep Learning Indaba in Dakar, and in 2025 gave a keynote at the Robotics and Automation in Africa workshop in Kigali. These activities put him back in conversation with communities building their own research capacity.

When Ashesi introduced its master’s in Intelligent Computing Systems, he welcomed the prospect of more applied research opportunities in Africa. His own experience gives that prospect substance: a faculty collaboration, time to pursue a capstone further, and opportunities to share the findings helped make research a possible future. RollE was designed to give more people a way into experimentation. For its creator, it did exactly that.

More News

Want to share a story?

We invite all members of the Ashesi community to share videos, photos, and story ideas. Contact the communications team at: website@ashesi.edu.gh

Activity Calendar

Featured Event: December 3, 2025

Christmas on the Hill
A festive end-of-year celebration featuring activities, music, and community bonding. This event brings together students, faculty, and staff to share in the holiday spirit before the break.