<p>The Network Intrusion Detection System (NIDS) is the core component of a digital network system. NIDS protects digital systems from threats such as malicious activities and unauthorised access. There are many sophisticated automatic intrusion detection systems whose operational concepts remain uninterpretable and secret. An explainable automatic intrusion detection system requires time to distinguish intrusion attempts from accidental anomalies. An explainable intrusion detection system also provides a way to generate essential justifications for cybersecurity forensics. This work, named BFO Optimised Explainable SVM with Fuzzy Feature Selection Variable Precision Rough Set for Intrusion Detection, abbreviated as BESFVID, is an attempt to bring a stable explainable intrusion detection system into practice. As a technical background, fuzzy feature selection and VPRS-based SVM are used in the BESFVID work. Two novel contributions, namely the Flexible Bacteria Forage Optimisation Algorithm and the Explainable Intrusion Detection Classifier, are introduced in the BESFVID work to achieve higher accuracy and precision in intrusion detection. It is also ensured that the overall intrusion detection time is controlled. BESFVID performance is evaluated using the OPNET network simulator with the NSL-KDD and UNSW-NB15 datasets.</p>