<?xml version="1.0" encoding="UTF-8"?>
<oembed>
 <version>1.0</version>
 <type>link</type>
 <provider_name>Rainbow Dash Network</provider_name>
 <provider_url>http://rainbowdash.net/</provider_url>
 <title>Cerulean Lulamoon-Spark (ceruleansparkold)'s status on Wednesday, 28-Sep-11 16:34:33 UTC</title>
 <author_name>Cerulean Lulamoon-Spark (ceruleansparkold)</author_name>
 <author_url>http://rainbowdash.net/ceruleansparkold</author_url>
 <url>http://rainbowdash.net/notice/561728</url>
 <html>@&lt;span class=&quot;vcard&quot;&gt;&lt;a href=&quot;http://rainbowdash.net/user/1235&quot; class=&quot;url&quot; title=&quot;Henrik Westerlund&quot;&gt;&lt;span class=&quot;fn nickname&quot;&gt;flaxx&lt;/span&gt;&lt;/a&gt;&lt;/span&gt; It wasn't actually so hard. The robot had 3 pressure sensors (Front left, front right and back). You just had to program it with a &amp;quot;response&amp;quot; to pressure on each sensor. The maze walls were all straight and all the turns were at predictable 90 degree angles, so the responses were typically some variation of &amp;quot;Stop, go back a little and turn a quarter or half circle, then go forward again&amp;quot;</html>
</oembed>
