Power & Utilities
Utilities transform by converting recurring inspection and event assessment into a persistent, software-orchestrated sensing layer.
- UK DNO investment
- £23.1B
- EU grid investment
- $70B+
- Restoration time saved
- ~30 min
- BVLOS target
- 2027
- UK DNOs investing £23.1B during RIIO-ED2
- EU grid investment exceeding $70B in 2025
- National Grid moving to centralised drone inspection
- UK CAA targeting routine BVLOS by 2027
Power and utilities is not one operating environment. It is a stack of distinct but connected asset classes: centralised generation, distributed generation, bulk transmission, local distribution, substations and converter stations, overhead lines, underground cables, offshore connections, hydro structures, wind and solar assets, control centres, telecoms, and field service operations. The asset base is geographically vast and operationally heterogeneous. This matters because physical AI succeeds
The strategic importance of this industry is rising. Electrification, renewable integration, resilience planning, and geopolitical energy security have pushed grid infrastructure toward the centre of industrial policy. The IEA has repeatedly argued that grids are becoming a critical bottleneck for secure energy transitions, while European investment needs are rising materially. That increases the importance of doing more with existing workforces and asset fleets, and of extracting better conditi
Utilities already run highly digital operational cores. Control rooms, protection systems, SCADA, outage management, energy management systems, and various planning environments are mature by comparison with many industrial sectors. But the field layer remains much less digital. Even sophisticated utilities still depend on scheduled patrols, periodic inspections, manual asset checks, photography, thermography, helicopter surveys, contractor rounds, and human interpretation of defect evidence. Th
Utilities transform by converting recurring inspection and event assessment into a persistent, software-orchestrated sensing layer.
- 01What changes operationally
Operationally, utilities move from episodic, manually gathered condition evidence toward higher-frequency, machine-captured evidence. Substation rounds become more structured. Drone inspection becomes less ad hoc and more scheduled. Fault finding bec
- 02What changes economically
Economically, the main shift is that the cost of “seeing the asset” falls. Today, the cost of seeing can include travel, foot patrols, shutdowns, access permits, helicopter flights, rope access, specialist inspections, or simply delay. Physical AI lo
- 03What changes strategically
Strategically, utilities begin to design for robotic access and data integration. That means better site connectivity, charging and docking points, mapped routes, asset identifiers, computer-readable inspection points, stronger data fabrics, and EAM/
- 04Which parts of the value chain are most affected first
The first parts of the value chain to change are those dominated by repeatable sensing rather than dexterous action: line and corridor inspection, substation inspections, hard-to-access generation checks, post-event condition assessments, and some as
- 05What remains hard to automate
What remains hard is everything that combines physical dexterity, live electrical risk, and open-ended judgment. This includes switching tasks, live-line work, emergency storm restoration in messy environments, ad hoc repairs, most heavy vegetation c
The rest of the 2026 Power & Utilities report
9 more sections — the workflow map, use-case atlas, technology stack, vendor landscape, economics and adoption timeline. Free account, and you also get the daily evidence digest.
- 01Workflow Map
- 02Value Today
- 03Use Case Atlas
- 04Technology Stack
- 05Vendor Ecosystem
- 06Economics & Business Case
- 07Risks & Readiness
- 08Adoption Timeline
- 09Frequently Asked Questions