SENSYN ROBOTICS Begins Full-Scale DJI Dock 3 Inspection Operations at Idemitsu Kosan’s Hokkaido Refinery

prtimes.jp reported on October 6, 2026, that SENSYN ROBOTICS had begun full-scale DJI Dock 3 inspection operations at Idemitsu Kosan’s Hokkaido Refinery. We have not been able to verify the primary source, so please refer to the original for details.

At sites such as Idemitsu Kosan’s Hokkaido Refinery, where patrols across a vast area coincide with stringent information security requirements, a key condition for deployment is not simply reducing staffing for inspections but being able to manage control, storage and approvals without using overseas cloud services. The significance of this operation is that SENSYN ROBOTICS has demonstrated a configuration that meets those conditions. The design also accumulates images as fixed-point observation data, a prerequisite for future AI anomaly detection, making this an effort to build a data platform rather than simply use drones. Idemitsu Kosan’s emphasis on a domestic data center environment illustrates the strict operational requirements at energy infrastructure sites.

What is new in this announcement is not replacing refinery inspections by introducing drones, but moving into full-scale operation a system linking DJI Dock 3, SENSYN CORE, a domestic data center and Datastore. The company had already conducted a demonstration using DJI Dock 2 in 2024, so it is reasonable to view this as a continuation of that effort, with an updated aircraft and operations becoming established. In other words, the issue has shifted from “Can a drone fly there?” to “Can inspections, storage and approvals all be managed domestically at a refinery with stringent security requirements?” This newspaper has no related coverage, so any connection to past reports is limited to the timeline in this announcement. The progression from verifying effectiveness in a 2024 demonstration to moving into full-scale operations in summer 2025 is a typical transition from proof of concept to integration into daily operations. What matters here is that the design explicitly avoids routing data through overseas cloud services. Idemitsu Kosan’s comments also cite the ability to control the aircraft and collect data in a domestic data center environment as a factor in the deployment decision, suggesting that information management constraints, rather than performance, are a key condition for adoption at the site. In terms of industry structure, this case goes beyond inspection equipment: it reconfigures input, processing, storage, approvals and future AI analysis as one operational chain. Traditional visual patrols required considerable travel time and placed a heavy burden on workers in hazardous areas. In this case, images are stored in Datastore immediately after capture and can be approved on a PC, making digitized inspection records part of the process. The design also accumulates fixed-point observation data that can serve as a training base for AI anomaly detection, giving the initiative a stronger focus on creating data assets than on one-off labor savings. One unresolved issue is how far integration with AI analysis will progress. The release outlines future plans for integration but does not specify which anomalies will be detected or at what accuracy. If operations at the Hokkaido Refinery are expanded to other sites, another key question will be whether the same security and network conditions can be replicated. The next point to watch is how widely the system can be deployed while keeping aircraft control, data storage and approval workflows entirely domestic.

At Japanese refineries and plants, deployment conditions may depend more on systems for domestic control and data management than on the aircraft itself. In this configuration, aircraft control at a domestic data center and record management through Datastore are central, increasing the importance of requirements for software and data center operations serving domestic infrastructure sites.