Rakuten Symphony has developed a prioritized list of the top 25 AI initiatives for communications service providers and divided them into five categories.
Site Selection Image Recognition is a number one priority for the Network Site Management and Roll Out category.
The process of selecting, evaluating and acquiring a cell site property is long and complicated. Once it gets to the point of starting construction, time is of the essence and AI can work with onsite field engineers to speed up the process.
At the start of site construction, the field engineer is responsible for receiving a lot of the technology and materials – from construction materials to equipment cabinets, to RAN systems and more. All of this equipment must be accounted for to keep the build out on track, and also added to inventory so that the MNO has a record of its equipment.
Currently, field engineers manually complete their checklists and upload images as a proof. When that work is validated by management there are often errors like incorrect serial numbers, incorrect photo upload, or incorrect tagging. This manual process takes a lot of time.
This manual process has become automated and AI-enhanced with Intelligent Site Audit features within the Rakuten Site Manager solution suite. Rakuten Site Manager is designed to accelerate network planning and deployment for MNOs and communications service providers. Rakuten Site Manager provides end-to-end visibility and automation into all network rollout-related tasks.
By using Rakuten Site Manager, the field engineer has a tool to complete their equipment installation task/checklist and upload the pictures of the installed equipment as proof. This checklist will be later submitted to the supervisor who will verify it based on the pictures attached to determine whether or not the installation, checklist/test case is correct.
The Intelligence Site Audit is an image recognition system integrated with the site survey checklist. When field engineers upload survey images, the AI module analyzes the images and compares them against predefined criteria.
Real-time feedback is provided to the engineers regarding any discrepancies detected, allowing for immediate corrective action. The model supports the following scenarios:
The use of Intelligence Site Audit improves accuracy and efficiency in field operations. It also offers reduced time and cost associated with rework and a long verification process by enabling real-time corrective actions. Real-time error messages are displayed to field engineers within seconds of image upload. The system achieves a minimum accuracy rate of 95% in detecting picture discrepancies.
