Robots are crawlers/bots/spiders/etc from search engines (Google, Yahoo!, Bing, and a host of others). They check for updated content on active sites like ours on a nearly constant basis.
As to how robots are identified.. each connection is supposed to include a "user agent" string identifying the client used to make the connection. For most of us, this "user agent" string identifies our web browser and its version. Our bbs software has an internal list of known user agents for many many robots that it uses to identify robot access.
"Online now" is actually "Online within the last X minutes", with X currently set to 60. The reason this is set so high is that many users complained that our XenForo bbs software's default of 15 minutes was causing the new messages list to reset too quickly for them.
Moving this thread to About TUG BBS.