Sector deep-dive
Where robots and autonomy are already at work across the energy system — from boiler walls and live transmission lines to solar fields, turbine blades, geothermal rigs, mines, data centers, and orbit. Curated, with a primary source on every entry.
Deployed robots working in the field today · Announced contracted or piloted, not yet operational · Potential early R&D or concept
No entries match your filters.
= also profiled in the Companies dataset. Every card is cited; primary sources preferred. As of 2026-06-16.
September 2026 policy update: DOE’s Genesis Mission Consortium lists robotics and automation as a working-group area. The application inventory above was last comprehensively checked June 16, 2026.
Energy-robotics companies featured above that aren't yet in the Companies dataset — a running shortlist of what to add next.
Which energy tasks justify a robotics trial, what prevents a purchase, and which problems require federal action? This unofficial DOE inquiry asks operators and suppliers for evidence that can select projects for the draft Robotics Executive Order.
Draft concept · Not an official DOE document
Project Prime Mover: Robotics for Inspection, Production, and Laboratory Work
Proposed purpose: identify tasks where robotics could reduce hazardous exposure, increase accepted production, or shorten experimental turnaround. Ask respondents to document one operating task or failed deployment before recommending a program. The proposed DOE lead would use responses to select mission trials, refer rule-specific barriers to the responsible regulator, and identify research gaps. An RFI response is neither an award application nor a promise of purchase. No submission channel is open for this unofficial draft.
Attach a task record to any deployment claim. Use a site alias in a public response. Separate laboratory demonstrations, paid pilots, routine operations, and repeat orders; identify teleoperation and vendor support in each.
Editorial proposal, not a DOE selection. These missions offer a named buying decision and measurable work. Respondents may propose a different task with equivalent evidence.
Tank, pipe, and vessel inspection. DOE documents robotic tank inspection at Hanford. Ask an asset owner for required defect detection, coverage, retrieval method, exposure, and cost per accepted inspection. First test: seeded defects in a representative mock-up. No field trial without a recovery plan.
Weld, finish, or tend machines for a defined part family. Start through MEP automation support. Ask the plant for drawing tolerances, changeover frequency, rework, integration hours, and cost per accepted part. Compare conventional automation with learned control; the buyer sets the acceptance test.
Transfer and prepare samples under a lab's approved protocol. Ask for traceability errors, contamination, human interventions, and reproducible results per shift. Start with nonhazardous surrogates. Retain manual fallback; faster handling matters only if the experiment remains valid.
Within 90 days after the response window closes, proposed: publish a task shortlist, an evidence-gap register, and referrals naming the responsible agency. For each shortlisted task identify a willing host, an acceptance authority, a funding/authority review, and the missing evidence before a trial. Publish reasons for deferring tasks. Missing reliability evidence goes to R&D; proven economics with a financing gap goes to lenders; a citation-specific barrier goes to its regulator.
Agency roles: propose DOE Environmental Management for cleanup inspection and DOE Office of Science for laboratory tasks, subject to office agreement and mission authority. NIST/MEP would support manufacturing assessment; NIST would support measurement. OSTP would coordinate cross-agency referrals. DOE does not control private plants, mines, utility procurement, or other agencies' waivers.
Engagement interest — a checkbox block on the response form
Beyond written responses, the inquiry proposes three tracks so the roadmap is built on observed evidence, not just filings:
Operators and workers review the same task record. Output: disputed baseline assumptions, staffing effects, and conditions for a trial.
Labs and independent evaluators turn proposed tasks into test protocols. Output: acceptance metrics, withheld test cases, data rights, and an estimated test cost.
At consenting sites: inspect logs, observe a full work cycle including recovery, and compare claims with the task record. A demonstration alone cannot establish long-run reliability.
Unofficial draft authored for this tracker as a policy concept — not a Department of Energy document and not affiliated with DOE. Drafted 2026-07-16; revised 2026-09-24. Proposed missions and deadlines are editorial recommendations, not announced programs.
Sourcing note: entries are confirmed against company case studies or reputable trade/press coverage as of August 2026. Deployment status (deployed / announced / potential) is an editorial read and evolves — follow each source for the current picture. Spotted a company we're missing? It probably belongs in the Companies tracker too.