Project Summary
Problem and Evolution
High-voltage direct current (HVDC) converter station valve halls generate strong and hazardous electromagnetic fields (EMF), making them inaccessible to personnel while in operation. Consequently, live inspections of assets are not possible. Instead, halls are accessible once per year during a scheduled maintenance outage, limiting the ability to diagnose issues in real-time, leaving a significant gap in monitoring of live assets, and increasing the risk of undetected faults, unplanned outages, and inefficiencies in network operation. As HVDC infrastructure expands rapidly to connect new offshore and onshore renewable generation, this lack of continuous operational insight is becoming a major constraint on efficient network operation and planning.
Addressing this challenge, SSEN Transmission (SSEN-T) launched the NIA AIM High Project, introducing an autonomous robot to perform monitoring tasks within an HVDC valve hall. The Project demonstrated safe robotic operation in high-EMF environments, providing new capability to collect continuous data from previously inaccessible areas. Following its success, SSEN-T will deploy seven robots as part of business-as-usual (BAU) HVDC operations in July 2026.
While AIM High demonstrated robotic feasibility, it also exposed a new challenge: the vast quantity of sensor and camera data generated daily requires efficient data management and automated analysis to efficiently handle large volumes of real-time data. Manual interpretation is labour-intensive and reactive, offering limited insight into operational efficiency or emerging faults.
The ODIN Discovery Phase, demonstrated the feasibility of automated interpretation and diagnostics in an operational environment, utilizing the autonomous robot, deployed at the Blackhillock HVDC converter station. The Project proved that artificial intelligence (AI) and statistical techniques could identify equipment, extract thermal features, and visualise operational trends. This confirmed that robotic data can be used to understand component performance dynamically rather than retrospectively, significantly reducing manual data processing.
Building on this foundation, the Alpha Phase will trial integration of a condition monitoring solution, enhancing the Artificial Intelligence Data Analytics (AIDA) platform; a cloud-based system developed by Ross Robotics that integrates robotic, thermal, and operational data to provide real-time asset insights, enabling proactive detection of potential faults and data-driven maintenance decisions. The Alpha Phase will extend AIDA’s capability to:
- Detect early signs of inefficiency or stress;
- Quantify efficiency losses;
- Generate evidence to support condition-based maintenance and asset investment decisions.
The Beta Phase (12-18 months) will trial a scalable, industrialised version of the solution across multiple HVDC sites, preparing for full BaU deployment.
Challenge
ODIN directly addresses Challenge 4, Aim 2 and Theme 1 by applying AI to analyse large volumes of sensor and camera data from HVDC assets. Through pattern recognition, anomaly detection, and trend analysis, ODIN identifies early signs of degradation in HVDC infrastructure, enabling intervention before faults occur and reducing unplanned outages.
By integrating robot sensor data with AIDA analytics, ODIN advances condition-based and predictive maintenance, ensuring targeted interventions, reduced downtime, and longer asset lifespans. As low-carbon technologies expand, ODIN’s continuous AI-driven monitoring helps manage increasing variability and operational stress, minimising losses and enhancing system resilience.
Users Needs
Primary users are UK Transmission Owners (TOs) responsible for HVDC converter systems operation and maintenance, while secondary users include Original Equipment Manufacturers (OEMS) and the engineering supply hain. Their needs have evolved from periodic inspection data toward continuous, interpretable intelligence that supports proactive decision-making.
ODIN provides real-time operational insight and quantified efficiency metrics, supporting data-driven asset management, enhanced workforce safety, and reduced environmental impact through fewer site visits and extended asset lifespans.
As the UK’s HVDC network expands (currently approximately 65 valve halls); ODIN’s scalable, interoperable approach offers significant benefits across the GB transmission system, supporting consumer value through reduced losses, improved reliability, and accelerated progress toward Net Zero.
Innovation Justification
ODIN delivers a transformative approach to monitoring and optimising High Voltage Direct Current (HVDC) converter systems, directly addressing Challenge 4; Aim 2 (reducing efficiency loss in the context of greater network utilisation and low-carbon deployment) and Theme 1 (innovation to improve efficiency of network operations).
Through the integration of autonomous robotics, artificial intelligence (AI), and cloud-based analytics, ODIN provides continuous, real-time insights into converter hall conditions that are otherwise inaccessible during live operation. The Project correlates thermal, acoustic, and optical data from robotic sensors with operational load and ambient parameters to identify efficiency losses caused by thermal imbalance, electrical stress, or component degradation. These insights enable proactive maintenance and performance optimisation, reducing downtime, cost, and carbon intensity across HVDC operations.
In doing so, ODIN provides a first-of-a-kind evidence-based link between real-time operating conditions and measurable efficiency in live HVDC environments—a key enabler of more transparent, optimised network operation.
Previous Innovation and Learning
Discovery Phase demonstrated that inspection data from the Blackhillock converter robot could be automatically interpreted using machine-learning algorithms. Components were identified, thermal patterns analysed, and visualisations generated to support maintenance teams. This confirmed that AI can deliver consistent, repeatable diagnostics and reduce manual workloads.
The Alpha Phase now builds upon this evidence by enhancing Ross Robotics’ AIDA platform to deliver end-to-end condition monitoring and diagnostic capability. This includes:
- Multi-source data integration (thermal, UV, acoustic, visual, and telemetry).
- Automated component trending and anomaly detection.
- Configurable alerting and dashboard visualisation.
- Predictive modelling of component degradation and efficiency loss.
Early prototype interfaces, shown in (Appendix_Q6_Figures_1.2-1.5), illustrate how operators will visualize inspection histories, component statistics and environmental status dashboards in real-time.
Lessons from AIM High and ODIN Discovery highlighted three technical priorities:
- Robust data pipelines to manage continuous sensor feeds;
- Validation methods for AI decisions to build operational trust;
- Clear data-governance protocols to ensure compliance.
ODIN Alpha embeds these through modular architecture, standardised data schemas, and explainable AI algorithms. Network engineers will be able to interrogate AI-driven outputs directly, fostering confidence and accountability.
Stakeholder Engagement
ODIN has been developed collaboratively with HVDC operation and maintenance engineers, data-science specialists and is openly collaborating with other European Transmission System Operators (TSOs) using similar robotic systems, including RTE, Elia, 50Hertz, and National Grid. Feedback from these stakeholders refined ODIN’s user interfaces and alert hierarchies, ensuring outputs align with operational practice. (Appendix_Q6 Figure_2.1) outlines milestones for prototype delivery, validation and knowledge dissemination, supporting future GB-wide learning.
Innovation and State-of-the-Art Comparison
Current practice relies on manual surveys during annual outages, capturing only static snapshots. ODIN replaces this with continuous, intelligent analytics combining mobile robotic sensing and AI diagnostics. Compared to these reactive methods, ODIN delivers earlier fault detection, improved asset utilisation and measurable reductions in efficiency loss. No existing system integrates autonomous data acquisition and live analytics within high-EMF HVDC environments, making ODIN a global first-of-kind innovation.
Readiness Levels
Please refer to figure 1 attached.
SIF Funding
Integrating robotics, AI and operational-technology systems in high-EMF environments is technically risky and beyond BaU or price-control budgets. SIF funding enables shared risk, cross-network learning and independent validation of benefits that will inform future investment decisions. The Project scale is proportionate—sufficient to validate algorithms, prove interoperability and quantify efficiency gains without premature commercial roll-out.
Counterfactuals
Alternative approaches—manual inspection, fixed sensors, vendor-specific monitoring or reliance on SCADA alarms—were evaluated and dismissed. These methods cannot deliver comprehensive, continuous or predictive insight. ODIN uniquely provides quantifiable reductions in efficiency loss and establishes a replicable, data-driven operating model for HVDC networks across Great Britain.
Impacts and Benefits
Financial - future reductions in the cost of operating the network
Under business-as-usual (BaU) conditions, HVDC converter stations are inspected during planned annual outages and maintained reactively when faults occur. This approach results in higher operational expenditure, reduced system availability, and greater exposure to unplanned outages that constrain renewable generation. Baseline metrics include system availability (%), frequency of forced outages, mean time to repair (MTTR), and annual maintenance costs.
ODIN introduces condition-based maintenance (CBM) using autonomous inspection and AI-driven analytics. Continuous monitoring enables early fault detection, targeted interventions, and optimised scheduling, reducing site visits, labour hours, and unplanned repair events while extending component lifespans. The solution also streamlines data management, replacing manual review with automated analysis through the AIDA platform.
Across SSEN Transmission’s HVDC sites, implementation of ODIN’s AIDA platform across seven autonomous robots is forecast to yield cumulative net benefits of approximately £0.46 million in the first ten years, rising to £20.5 million over the lifetime of the assets. There are currently an estimated 65 HVDC valve halls across the GB transmission network. Successful implementation of ODIN across these assets would deliver material operational cost reductions, reliability improvements, and indirect carbon savings at national scale.
The quantified benefits, supported by a detailed cost-benefit analysis (CBA), underline ODIN’s potential to deliver significant long-term savings and operational resilience.
New to Market – products, processes and services
ODIN delivers three inter-related innovations that are new to the market:
- Product – the UK’s first untethered autonomous robotic sensor platform capable of operating safely within live HVDC valve halls demonstration.
- Process – a new methodology for data-driven, condition-based maintenance that replaces periodic manual inspection with continuous digital monitoring.
- Service – the AIDA cloud analytics platform providing AI-driven diagnostics, trend visualisation, and efficiency indicators accessible to operators in real time.
Collectively these enable a step-change in asset-management practice, offering GB network operators scalable, interoperable tools for predictive maintenance and efficiency optimisation. Discovery-phase engagement with engineers and TSOs validated user requirements for explainable AI and standardised data formats, ensuring that Alpha-phase outputs will be directly applicable to future BaU adoption.
Revenues - creation of new revenue streams
ODIN provides an opportunity for UK-based Ross Robotics’ long-term strategic growth and market leadership, scaling their workforce (50+new jobs), increasing revenue growth (£70million) by 2030, developing their supply chain and positioning them as a leading innovator in the autonomous robotics sector both domestically and internationally.
Environmental - carbon reduction, indirect CO2 savings per annum
The current inspection regime requires engineers to travel frequently to remote converter stations for data interrogation and investigations and depends on energy-intensive standby operation during outages. ODIN enables continuous monitoring with minimal site visits, reducing both vehicle emissions and energy losses from inefficient converter performance.
Others – not SIF specific
ODIN provides broader public and system benefits:
- Improved reliability and network resilience, reducing the risk of renewable curtailment.
- Enhanced safety, eliminating the need for personnel in high-EMF environments.
- Skills and Supply Chain, develops UK expertise in robotics, AI, and digitalisation, creating sustainable employment.
- Leading Innovation world-wide, ODIN is a global first-of-a-kind.
The Discovery Phase produced a Cost Benefit Analysis demonstrating a positive benefit–cost ratio for Alpha progression. The Alpha Phase will refine these metrics through live trials, with network-level consumer benefits reported using Ofgem and Innovate UK’s CBA framework and verified by SSEN Transmission’s asset-performance team.
