Project-manage an RSS submission against its real calendar. Anchor every date to the current edition's pages; the 2026 dates below are a verified historical snapshot (source map rows 1-3, checked 2026-07-08), not a promise about 2027.
rss-camera-ready, rss-artifact-evaluation).RSS shares its authors with three siblings, so choosing RSS is also choosing a slot in an annual rotation. Approximate recent geometry (verify each venue's live pages):
| Venue | Typical deadline | Typical event | What slipping past RSS costs |
|---|---|---|---|
| RSS | late January | July | Next comparable single-track slot is a full year away |
| IROS | ~March | ~October | Natural catch for a paper 6 weeks short of RSS-ready |
| CoRL | ~June | ~November | Right jury if the learning story dominates |
| ICRA | ~September | ~May/June | Broadest program; absorbs systems breadth |
The practical use: a project that misses the January bar honestly should aim at IROS or CoRL rather than submit thin — RSS's selectivity makes weak submissions expensive in reviewer goodwill.
| Weeks out | Milestone for a claim-plus-hardware paper |
|---|---|
| 12+ | Scientific claim fixed in one sentence; evidence plan sized to it |
| 10 | Hardware trial campaign running; failure attribution logged from day one |
| 6 | Ablations that isolate the mechanism complete |
| 4 | Full draft; supplement storyboard; internal red-team read |
| 2 | Anonymity sweep of PDF and media; OpenReview metadata entered |
| 1 | Abstract registered; buffer for robot downtime already spent |
| 0 | Paper in; supplement due one week later — do not design it that week, only pack it |
The one-week paper→supplement gap (2026: Jan 30 → Feb 6) is packing time, not filming time; footage shot after the deadline cannot support claims already frozen in the PDF.
For teams presenting at the current edition, the final week has its own checklist:
rss-artifact-evaluation).Nothing about 2027 was posted at check time. Watch these, in this order, starting early autumn 2026:
RSS/2027 group appearing.rss-author-response).Robotics submissions fail at handoffs; name owners early:
| Workstream | Single owner for |
|---|---|
| Claim custody | The one-sentence claim; vetoes scope drift in any section |
| Trial campaign | Protocol document, ledger integrity, daily campaign log |
| Media | Footage banking from every session; the supplement cut |
| Anonymity | The sweep across PDF, video, archive, and OpenReview fields |
| Deadline mechanics | AoE conversions, form fields, upload confirmations |
Footage banking deserves emphasis: clips must be captured during the campaign, because the supplement freezes one week after the paper and reshoots after the paper deadline cannot support frozen claims.
[Cycle position] planning 2027 / drafting / submitted / rebuttal window / accepted / conference week
[Next hard date] <date + source URL, or 待核实>
[Critical path] <three tasks gating the next stage>
[Fallback venue] IROS / CoRL / ICRA + trigger condition
[Calendar risks] <robot downtime / visa / footage / rebuttal availability>