Unmanned and minimum manned offshore platforms are recently getting much attention due to their cost-saving potentials. They are expected to maintain continuous production for 24/7 while being operated from onshore control rooms. Such status is closely linked to the achievement of near zero- downtime performance which would require cost-effective maintenance strategies and data-dependent decision support system. Traditional corrective and scheduled maintenance practices may not be adequate in that regard and predictive maintenance for decision support may add new values. In the era of digitalization, like many other industries, huge amount of data are being collected in o&g sector that are also offering opportunities to move towards predictive maintenance. In this paper, we consider ultrafiltration membranes used to pretreat seawater in water injection process that are subject to degradation and required replacement in 3-5 years. It is believed that, due to associated lead time and lack of redundancy, predictive support for replacement may bring additional values. Data from an offshore platform in North Sea is collected and analyzed to identify prognostics and maintenance optimization options. A possible condition indicator for prognostics model development for replacement decision and a possible optimization scope for optimal chemical wash sequence is discussed constrained under existing limitations and challenges.