Object detection for Red Palm Weevil (RPW) using deep learning (DL) is an effective and innovative method to combat the infestation of this destructive pest in palm plantations and agricultural areas. RPW is a significant risk to palm trees worldwide, leading to environmental damage and significant economic losses. It is a highly destructive pest that infests different palm species, causing the death of the trees if left unchecked. Our approach integrates computer vision (CV), cloud computing (CC), DL, Internet of Things (IoT), and geospatial data to accurately detect and classify palm trees infested with RPW. DL has revolutionized CV tasks, together with object detection, by enabling machines to automatically learn complex patterns and features from enormous quantities of data. The conventional way to object detection includes classifiers and handcrafted features, which can be less accurate and cumbersome in handling complicated scenarios such as RPW detection. The study proposes an automated RPW detection and classification utilizing Chaotic Harris Hawks optimization with deep learning (ARPWDC-CHHODL) method in cloud environment. The ARPWDC-CHHODL model involves the concept of DL-based object detection and classification with a hyperparameter tuning strategy. In proposed model, the EfficientDet-D0 model is applied as an object detector, which comprises three major components such as EfficientNet as a backbone network, bi-directional feature fusion, and classification. To enhance performance of EfficientNet approach, Chaotic Harris Hawks optimization (CHHO) methodology used for hyperparameter tuning procedure. Besides, long short-term memory (LSTM) approach executed for detection of RPW. The simulation value of ARPWDC-CHHODL approach verified on RPW database and the outcomes can be inspected in terms of various measures. The obtained outcomes displayed the promising solution of ARPWDC-CHHODL model with recent methods of RPW detection methods.