Gazprom Neft is to work with Russia’s Innopolis University on developing a minimum-viable-product (MVP) pilot version of a navigation system for controlling unmanned specialist vehicles and equipment by late 2021.* This service will facilitate the future off-road usage of unmanned bulldozers, excavators and other machinery in building oil and gas field infrastructure, as well as in geological investigations. Deploying robot technology will deliver significant cost reductions in setting up new oilfields, in the future.
The navigator’s algorithms will utilise digital terrain maps to work out the optimum routes in operating specialist equipment and vehicles, in line with lidar (light detection and radar) survey data and drone video analytics.** A high level of detail will allow every individual tree or obstacle to be positioned in a given space, together with information on its height and thickness — all of which means vehicle and equipment routes can be plotted even in forested areas difficult to access without damaging the environment.
Deep-learning neural networks are used in operating the navigator, with algorithms analysing digital maps of an area and coming up with the best routes for vehicles and equipment, in line with their type and size. Engineering operators then feed the resulting recommendations into an application, which then transmits commands to these robotic vehicles’ onboard computers.
“Our company is working through more than 250 scenarios for using robots at oil fields, at production facilities, and in logistics. A large proportion of these scenarios, moreover, envisage moving robotic systems in environments with no roads, in difficult terrain, or within production facilities. Developing a specialist-equipment navigator will address the task of navigating robotic solutions and unmanned vehicles in line with the specific problems of these conditions.”Mikhail Korolkov Head of Digital Technology Centre, Gazprom Neft
“In terms of business benefits, using robotics mainly means work will be done faster, and to a better quality. That doesn’t mean savings in terms of operators’ pay, but potential reductions in construction costs as a result of increased efficiency. Addressing this task is important because developing a system like this means we can mark a start on the most important thing — moving machinery in automatic mode. Introducing these systems is going to happen gradually, with operators being involved in the initial stages.”Roman Fedorenko Project Manager, Innopolis University
The Innopolis University team developed a realistic robotic simulator to test the navigator, in which they created a virtual construction site, including a quarry, unpaved roads and grassed areas. The simulator replicates the processes of a UAV or drone, which gathers information to build digital maps and to reflect a bulldozer’s movements in independent operation.
* MVP — a “minimum viable product” is a pilot version of a digital service, with basic capabilities and functionality, ready for real-world testing.
** Lidar — “light detection and ranging” is a technology for acquiring and processing information about remote objects using active optical systems that exploit light absorption and diffusion in optically transparent media.