Sludge dewatering centrifuges usually operate using fixed control setpoints, performance is assessed through grab sampling of cake and centrate. This provides limited visibility of performance restricting the ability of operators or control systems to respond to changing sludge characteristics in real time. We describe an optimisation programme undertaken at a European wastewater treatment facility in collaboration with a major consulting engineer and machine-learning technology provider. Initial work established a statistical baseline linking feed sludge quality, polymer dose, centrifuge operating conditions and dewatering performance. A machine-learning control layer was introduced to optimise operating parameters. The final phase integrated the SOLiSENZ CSZ in-line centrate total suspended solids (TSS) sensor, providing continuous measurement of centrate solids, enabling closed-loop feedback of dewatering performance. Sensor measurements demonstrated excellent agreement with laboratory results. The TSS measurement transformed process visibility. The combined programme increased cake dry solids from 22% to 24%, reduced polymer consumption by 14%, and delivered a 19% reduction in overall chemical costs. Improved solids capture also reduced solids recycle loads to secondary treatment processes, while higher cake solids decreased sludge volumes, transport requirements and associated carbon emissions. The presentation will discuss the role of continuous centrate solids monitoring in enabling advanced dewatering control, together with the validation methodology, control architecture and quantified operational benefits achieved.