AMBA & BGA EXCELLENCE AWARD WINNERS
negatives. Operators will not tolerate constant false alarms, but from a safety perspective you never want to miss a real event. Balancing these two points is, therefore, a constant calibration exercise involving technology, user interface and staff training. Furthermore, privacy and trust around dealing with videos of families and children is critical, demanding strong controls over data collection, storage, access and retention, as well as clear communication about what is and isn’t being done with the footage. “A further consideration is environment variability. No two pools or beaches are the same. Lighting, reflections, water colour, camera positioning and swimmer behaviour can all differ. This means an AI model that performs well in one facility may still need careful tuning in another. Additionally, even a very accurate model can fail in practice if it doesn’t fit into existing lifeguard routines, alarm systems and reporting processes. A lot of our work relates to workflow and integration into operations, not just algorithms.” In how many settings and locations is Deep Blue currently in use and what’s next for it in terms of future development? Luke Cunningham: “Deep Blue is already operating in real‑world pool environments, ranging from family leisure pools to more complex multi-pool sites, with additional deployments in the pipeline. We are also operating in pilot mode on a few beaches. At this stage, we tend not to publish exact site counts, but the footprint is growing as we prove reliability in different types of facilities. “Looking ahead, there are three main areas of development. First, deeper integration with existing pool management and alarm systems so that operators experience Deep Blue as part of their standard infrastructure, not a standalone gadget. Second, more advanced analytics for safety managers, turning raw alerts into insights about staffing, layout and training. Third, geographic expansion, particularly within the Middle East, in partnership with large operators and regulators who see water safety as a strategic priority.” Have you been able to measure the impact of Deep Blue in a setting where it has been used? Luke Cunningham : “In drowning prevention, the ideal outcome is that nothing ever happens, which is statistically hard to prove. So, rather than waiting for a before-and-after incident count, we work with operators to track metrics such as time from the onset of risky behaviour to detection by the system; time from alert to lifeguard intervention, especially during drills; frequency and characteristics of near-miss events; and false-alert rates and how quickly staff can clear them. “Feedback from partners so far suggests that Deep Blue helps standardise and speed up responses in drills and that the records of incidents and near-misses it generates are valuable for training and internal reviews. Over time, as deployments scale, we expect
to build a stronger statistical picture, but operators are already seeing operational benefits in their day-to-day practice.” Are there any other applications of Deep Blue’s AI model, beyond drowning detection? Luke Cunningham: “The core capability we’ve built is real‑time understanding of human behaviour and risk around water. That naturally lends itself to adjacent applications, such as detecting unsafe behaviour around a pool deck, or people entering restricted areas, as well as monitoring after-hours or unsupervised use of facilities.” What can business schools do to better support student and graduate entrepreneurs? Karl Baz : “Designing courses and assessments so they can be applied directly to a student’s own venture makes a noticeable difference. Bradford does this well: many of my submissions have been built around real-life problems, which means the work is immediately useful, not just graded and forgotten. “It’s also important to create structured access to practitioners, founders, investors and industry experts who can give uncensored feedback. Academic perspectives are important, but they need to be complemented by people who have navigated fundraising, regulations and hiring. “Lastly, schools should strengthen the bridge between the classroom and the market. This enables clearer routes to pilot projects with corporate partners, exposure to early-stage investors and support for things like governance and legal setup. Fundamentally, student ventures should be treated as serious, long-term clients of the school and not as side projects.” What advice would you give to an aspiring entrepreneur who is considering an MBA or EMBA to get their business off the ground? Karl Baz: “Be very clear about the job you want the degree to do for you. An MBA or EMBA will not automatically produce funding or customers, or level up your career. What it can offer is a set of tools, a network and a structured way to think about decisions you will have to make anyway. “If your instinct is to learn by doing, as it is for me, then you can think of the degree as an amplifier, rather than a prerequisite. Ideally, you bring a live idea or venture into the programme and use every assignment to stress-test it. “The key questions to ask yourself are whether you will use the programme to work on your own business, rather than on case studies; if you can realistically handle the time and energy requirements while building something; and if you value the frameworks and network on offer through the programme enough to justify its financial cost. If the answer to these questions is ‘yes’, an MBA or EMBA can be a force multiplier. If not, you may be better off starting small and experimenting in the market, before coming back to formal study when you know exactly which gaps you want to fill.”
Ambition • ISSUE 6 • 2025 33
Made with FlippingBook - Share PDF online