<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Safety on DNTD Dynamics</title>
    <link>https://dntddynamics.com/tags/safety/</link>
    <description>Recent content in Safety on DNTD Dynamics</description>
    <image>
      <title>DNTD Dynamics</title>
      <url>https://dntddynamics.com/images/og-card.png</url>
      <link>https://dntddynamics.com/images/og-card.png</link>
    </image>
    <generator>Hugo -- 0.124.0</generator>
    <language>en</language>
    <lastBuildDate>Thu, 21 May 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://dntddynamics.com/tags/safety/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Micro-doppler classifier — not everything that moves is a person</title>
      <link>https://dntddynamics.com/lab-notes/mmwave-phase6-microdoppler-classifier/</link>
      <pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate>
      <guid>https://dntddynamics.com/lab-notes/mmwave-phase6-microdoppler-classifier/</guid>
      <description>Phase 6 adds a DBSCAN cluster builder and a rule-based micro-doppler classifier to the pipeline. Novel returns are now labeled PERSON, OBJECT, or UNKNOWN before they reach the zone logic. UNKNOWN passes through as PERSON — the classifier can never make the system less safe than the baseline.</description>
    </item>
    <item>
      <title>Persistent background map — the sensor shouldn&#39;t relearn the room every boot</title>
      <link>https://dntddynamics.com/lab-notes/mmwave-phase5-persistent-background/</link>
      <pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate>
      <guid>https://dntddynamics.com/lab-notes/mmwave-phase5-persistent-background/</guid>
      <description>The 15-second background learning phase is necessary — but only once. Phase 5 adds a persistent voxel map saved to disk after the first learn, reloaded on every subsequent boot. Also: a novelty-aware refresh gate that means a stationary person can never be silently absorbed into the background.</description>
    </item>
    <item>
      <title>Swept-volume workspace clipper — the sensor finally knows what the arm can reach</title>
      <link>https://dntddynamics.com/lab-notes/mmwave-phase7-swept-volume-clipper/</link>
      <pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate>
      <guid>https://dntddynamics.com/lab-notes/mmwave-phase7-swept-volume-clipper/</guid>
      <description>Phase 7 adds a swept-volume workspace clipper to the pipeline. Detections outside the arm&amp;rsquo;s reachable envelope are suppressed before they reach zone logic. This is the primary technical differentiator over fixed-mount sensors — and the reason the hardware configuration locked at 3× IWR6843AOP on the forearm link.</description>
    </item>
    <item>
      <title>Full safety stack live — background learning, ego-motion compensation, and the motionless person problem</title>
      <link>https://dntddynamics.com/lab-notes/mmwave-full-safety-stack-live/</link>
      <pubDate>Sat, 16 May 2026 00:00:00 +0000</pubDate>
      <guid>https://dntddynamics.com/lab-notes/mmwave-full-safety-stack-live/</guid>
      <description>The IWR6843AOP is now running a complete ROS 2 safety pipeline on the Jetson — CLEAR/CAUTION/STOP zone detection with ego-motion compensation, voxel background learning, fault handling, and a heartbeat watchdog. One non-obvious bug dominated the session: velocity filtering was silently eating motionless people.</description>
    </item>
  </channel>
</rss>
