L
Admin
[Discussion prompt / news context, rfforum Editorial Team]
Late October 2026 saw fresh 60 GHz connectivity announcements aimed at high-bandwidth, low-latency local links (e.g. Movandi MV6055-class platforms covered on Microwave Journal and everything RF news pages: https://www.microwavejournal.com/articles/topic/3369-rf-microwave-industry-news , https://www.everythingrf.com/news). Setting vendors aside, what should RF engineers ask before a factory-floor or robot-cell pilot?
1. Link budget is not the hard part alone
Oxygen absorption and blockage matter, but so do multipath from metal racks, human body shadowing, and whether the MAC assumes LOS. A 5 Gbps demo in a quiet lab does not transfer to a welded-steel cell without a fade margin plan.
2. Antenna and beam management
Phased arrays help, but scan loss, radome/paint, and mounting height change the usable SNR. Ask for EIRP masks, beam codebook limits, and what happens when two cells interfere on the same unlicensed channel.
3. Latency and isochrony
"Physical AI" and closed-loop control care about tail latency and sync, not only peak throughput. Ask for latency histograms under load, not a single best-case number.
4. Coexistence and compliance
Unlicensed 60 GHz rules, radar sensing in the same band, and industrial EMC all constrain duty cycle and power. Confirm the regulatory region and the coexistence test plan early.
Open questions for the room:
- Have you deployed or evaluated 60 GHz (or 57–71 GHz) inside a metal-rich factory / warehouse, and what was the binding limit — blockage, interference, or thermal/PA backoff?
- Do you treat 60 GHz as a cable replacement for a known path, or as a mesh that must survive people walking through the beam?
— rfforum 编辑部 / Editorial Team
Late October 2026 saw fresh 60 GHz connectivity announcements aimed at high-bandwidth, low-latency local links (e.g. Movandi MV6055-class platforms covered on Microwave Journal and everything RF news pages: https://www.microwavejournal.com/articles/topic/3369-rf-microwave-industry-news , https://www.everythingrf.com/news). Setting vendors aside, what should RF engineers ask before a factory-floor or robot-cell pilot?
1. Link budget is not the hard part alone
Oxygen absorption and blockage matter, but so do multipath from metal racks, human body shadowing, and whether the MAC assumes LOS. A 5 Gbps demo in a quiet lab does not transfer to a welded-steel cell without a fade margin plan.
2. Antenna and beam management
Phased arrays help, but scan loss, radome/paint, and mounting height change the usable SNR. Ask for EIRP masks, beam codebook limits, and what happens when two cells interfere on the same unlicensed channel.
3. Latency and isochrony
"Physical AI" and closed-loop control care about tail latency and sync, not only peak throughput. Ask for latency histograms under load, not a single best-case number.
4. Coexistence and compliance
Unlicensed 60 GHz rules, radar sensing in the same band, and industrial EMC all constrain duty cycle and power. Confirm the regulatory region and the coexistence test plan early.
Open questions for the room:
- Have you deployed or evaluated 60 GHz (or 57–71 GHz) inside a metal-rich factory / warehouse, and what was the binding limit — blockage, interference, or thermal/PA backoff?
- Do you treat 60 GHz as a cable replacement for a known path, or as a mesh that must survive people walking through the beam?
— rfforum 编辑部 / Editorial Team
