Case Study

Saipem improves safety and operational efficiency with autonomous industrial inspections

Saipem has adopted a platform for managing autonomous robots, making industrial inspections safer, repeatable, and scalable through the use of quadrupeds such as Unitree's Go2.

THE challenge

Manage inspection missions in large, unstructured, and potentially hazardous industrial environments, overcoming accessibility, connectivity, and management limits of large volumes of visual and spatial data, to make operations standardized and replicable.

SCENARIO

Bringing autonomous robotics to complex industrial inspections

Industrial inspection activities are becoming increasingly complex in contexts where assets are distributed across extensive sites, operational continuity is critical, and many areas are difficult or risky for operators to access. In this context, Saipem, a global leader in engineering and the execution of large projects for the energy and infrastructure sectors, needed to evolve its inspection model, combining safety, efficiency, and repeatability.

The goal was not simply to introduce robotics in the field, but to do so in an integrated, scalable manner consistent with the constraints of industrial environments, often unstructured, with limited connectivity and high requirements in terms of cybersecurity and data protection. With the support of Roboverse Reply, a platform for Autonomous Robot Management was designed and implemented to orchestrate robotic missions, collect field data, and generate a digital representation of the inspected environment, promoting closer integration between physical assets and digital processes.

the solution

Cloud and Infrastructure as Code: a cutting-edge middleware

The new Autonomous Robot Management platform integrates the quadruped robot Unitree Go2, equipped with a 3D laser scanner, onboard computer, and custom payload, and allows operators to configure and launch autonomous missions through an intuitive graphical interface, without requiring specialized robotic skills.

The robot autonomously navigates the environment, builds and updates a three-dimensional map of the site in real-time, and collects visual and spatial data along the way. Inspection data is stored locally on the robot and automatically synchronized with the central system as soon as connectivity is available, ensuring operational continuity even in completely disconnected environments. Images captured during missions undergo automatic anonymization through computer vision, ensuring compliance with privacy regulations.

The platform is built on an open architecture, compatible with various types of robots and drones, and supports centralized deployment across multiple sites. A physics-based PhysX simulation environment and RTX rendering allow for testing and validating new use cases in a virtual replica of the operational environment before physical implementation.

The results

Less risk, more data, and continuous operations: a new inspection model for Saipem

By adopting the new platform, Saipem has created the conditions to evolve the inspection model into a safer, continuous, and digital approach. The main results achieved include both daily operations and the ability to extend the model to new application scenarios:

Greater security
for the staff

Inspection activities in hazardous or hard-to-reach environments are now conducted by robots, limiting the direct exposure of operators.

Greater efficiency
perspective

The autonomous execution mode of missions and the Touch & Go interface reduce setup and launch times, increasing the frequency of inspections without increasing the operational load.

Quality and consistency
of the data

The availability of data localization and standardized collection protocols ensure accurate, repeatable, and comparable outputs over time, regardless of the operator or site.

Continuity
operational

Offline-first architecture removes dependency on network connectivity, making the platform reliable even in the most isolated environments.

Scalability
.

The platform is ready to expand to new sites, new types of robots, and new use cases, including the integration of AI models for automatic anomaly detection and predictive maintenance.

Saipem is an international group specialized in engineering and construction services for the energy sector. With a presence in over 60 countries and a portfolio ranging from upstream oil & gas to infrastructure for renewable energies, Saipem operates in some of the most complex and remote environments in the world. The company continuously invests in technological innovation to improve the safety, efficiency, and sustainability of its operations.

Roboverse Reply

Roboverse Reply is the company of the Reply group specialized in advanced robotics solutions and autonomous systems for industrial applications. Roboverse Reply designs and develops hardware-software platforms for the management of mobile robots, drones, and heterogeneous fleets, with an approach focused on open integration, security, and scalability. It operates in sectors such as energy, manufacturing, infrastructure, and logistics, developing tailored solutions that combine autonomous navigation, digital twin, artificial intelligence, and advanced sensing technologies.