At the recent Lagos Mobility Innovation Summit, a quiet session on engineering workforce planning became one of the hardest to leave. While most attention went to electric buses, fare systems and intelligent transport networks, visitors stayed to hear about StaffLens Beta, a new AI powered platform that helps transport operators forecast and allocate engineering teams more systematically.

StaffLens Beta is built around a forecasting engine that comes from published research and a doctoral thesis on machine learning for transportation workforce planning. Instead of treating staffing as an afterthought behind buses, tracks or ticketing equipment, the tool puts engineering talent at the center. It helps agencies estimate how many design engineers, systems specialists, schedulers and maintenance planners they will need on upcoming projects, even when the data is incomplete or inconsistent.

The research that inspired the platform began from an uncomfortable reality. Many transit organizations in Nigeria and elsewhere do not have clean, long-term datasets. They rely on scattered HR files, partial project logs and a few traffic or reliability indicators. Poor engineering staffing quietly fuels delays, redesigns and burnout, yet rarely appears in public debates. The underlying work behind StaffLens Beta focused on this neglected problem instead of more popular topics such as route optimization or passenger demand prediction.

To address that gap, the framework blends operations research with modern machine learning under a strict constraint: it must work with small, messy datasets. For many Nigerian operators, usable project records may only go back a few years, traffic counts may be manual and staffing sheets may live in inconsistent spreadsheets. The research embraced that reality and built methods that tolerate noise rather than assuming clean data. That practical choice helped StaffLens Beta stand out at the summit.

A Lagos based technology company then translated the academic methods into software. The forecasting and simulation algorithms sit inside a web interface, while collaboration tools, dashboards and permissions wrap around them. What appears in journal articles as notation and equations shows up for users as sliders, charts and scenario buttons that project managers can understand without a research background.

The forecasting engine powers the core module that estimates engineering workforce needs across projects, regions and time horizons. It flags likely staffing gaps, highlights periods of overload and lets users test scenarios such as hiring, retraining or outsourcing before they commit. Other parts of the platform handle project tracking and budgeting. At the time of the summit, StaffLens Beta was still in pilot use with a handful of early adopters, with wider rollout planned after additional testing.

Across early pilots, organisations using StaffLens Beta reported measurable gains: about a 27 percent reduction in last minute reshuffling of project engineers, an 18 to 22 percent improvement in forecasting reliability even with sparse data, and roughly a 15 percent cut in internal planning time as the platform’s use expanded beyond engineering workforce planning into wider operations.

A packed panel titled “From Papers to Platforms: Making Transport Research Implementable” examined how StaffLens Beta came to exist. A university professor, a senior engineer from LAMATA, a union representative and an early adopter discussed the platform as a rare case of focused academic work moving into daily practice. Much of the discussion revolved around the research mindset behind the tool and the broader issue of why so many strong ideas fail to reach industry.

Halfway through the session, the moderator revealed that the research underpinning StaffLens Beta was authored by Ogungbire Adedolapo, a Nigerian transportation machine learning researcher, Tau Beta Pi fellow and recipient of a distinguished doctoral fellow in the United States. The panel noted how unusual that level of recognition is for Nigerian researchers and welcomed the fact that the work is already influencing operations at home.

The moderator also noted that Ogungbire’s other research works are finding practical use. One ongoing effort is exploring how his work on vision-based machine learning can be adapted to Nigerian roads to reduce motorcycle crashes and fatalities in Lagos and across the country. Earlier this year, a team of AI researchers used his published vision techniques to win a local safety competition on real time helmet enforcement. Their prototype combined an ensemble of deep learning vision models with a robust, Nigeria specific data pipeline and targeted data augmentation steps to achieve highly accurate detection.

For the professor on the panel, the central lesson was the decision to work on overlooked but relevant problems. The government engineer, Mr. Bode, stressed that because the methods were published openly rather than hidden inside a proprietary system, agencies and operators can study and adapt them. The union representative, Mr. Patrick, added that better forecasting of engineering workload supports healthier staffing negotiations and can help reduce burnout among members.

The conversation eventually moved beyond transport. Experts from health planning, power utilities and construction pointed out that they face similar challenges in forecasting scarce technical staff and said they hoped to adapt the same implementation first mindset. A quick poll after the session showed that more than 80 percent of the Nigerian researchers in the room saw the underlying work as influential and believed they could apply similar approaches in their own sectors.

By the close of Lagos Mobility Innovation Week, “like StaffLens Beta” had become shorthand for research that does not stay on the shelf. Startup founders compared notes on neglected areas such as safety inspections and maintenance staging, while younger researchers left thinking about how to design projects that can survive outside the lab and inside real organizations