What the research shows about remote guidance and AI trained on your own protocols — and what it means for aviation, energy, and healthcare teams.
Aviation, energy, healthcare, and manufacturing all run into the same wall: there aren’t many people who can do the hard version of the job well, moving them around is expensive, and the ones who can do it are getting older, not younger.
For a long time, the only fix was to physically move people. Fly the expert to the site. Fly the trainee to the expert. Wait until someone qualified is free. That constraint — not the cost of any particular technology — is what actually drives training and support costs in high-consequence work.
Two things are changing that now, and they work differently:
Both get expertise to the job without getting the expert on a plane.
Surgery is the most studied version of this, because surgeons have been mentoring each other remotely for years and someone finally wrote it down.
SurgeryA review covering 66 studies, mostly general surgery and urology, found remote mentoring held up. Outcomes were comparable to in-person mentoring, and patient safety didn’t get worse in any of them. That’s a fair basis for treating remote mentoring as a real substitute for physical presence — not a compromise.
Audio alone is bad at this job. The platforms that perform well let the remote expert draw directly onto the operating field in three dimensions. Tested on first responders performing a simulated emergency airway procedure, AR-based visual guidance beat audio-only on both speed and technique.
AviationThe U.S. Air Force ran the same test on aircraft maintenance. Technicians using an AR platform made 53% fewer errors and discrepancies. Technicians relying on traditional methods installed parts incorrectly 57% more often. Task time was roughly the same — the accuracy gain didn’t cost anyone speed.
The same setup already runs commercially for aircraft-on-ground situations: a remote expert walks a field engineer through a fix over live video instead of everyone waiting for a specialist to show up.
Energy & UtilitiesLinemen and substation technicians are doing the same thing with smart glasses. Hit something unfamiliar, start a video call, the remote expert sees exactly what the technician sees, and the fix happens on the spot instead of after a truck roll.
For work at height or near live equipment, that also means fewer people standing in the dangerous spot to begin with.
A phone call tells someone what to do. A shared video feed lets the expert show them, at the exact spot, in real time. That’s the difference between remote advice and remote expertise.
The usual fix has been documentation: write down the procedure before the expert leaves. It’s never really worked, because the useful part of expertise isn’t the procedure. It’s the judgment call, the workaround nobody wrote down, the reason a tolerance is set where it is. Manuals capture the formal steps and miss almost everything else.
A system trained specifically on an organization’s own material can do something a manual can’t: hold onto the reasoning behind a decision, not just the decision. Recent research found that conversational AI can lower the cost and skill barrier to documenting a process properly — so organizations can keep what they know even as the people who know it move on.
Learning by doing puts a skill in someone’s hands before the moment they need it. Remote expertise puts the expert in the room — a person or a trained system — once that moment has already started. Neither replaces the other. The strongest training programs in aviation, energy, and healthcare already run both.
Built on Apple Vision Pro, Tandem lets a remote expert see exactly what a field technician sees and draw directly into their field of view, hands-free, while the work is happening. Same mechanism as the surgical and aviation research above — already running in manufacturing, field service, logistics, oil and gas, and aviation.
Session Recording is where this connects to the second half. Every recorded session is a record of how an expert actually solved something — the same knowledge the workforce data says organizations are losing. The roadmap points this direction, with AI agents and model customization planned alongside the live-expert product.
Live Session — Tandem
Findings are drawn from peer-reviewed studies, industry research, and case reporting identified through a review of published literature.
Content