If you are happy for (want) all devices to be able to talk to each other on a single LAN, then you do not need to have different actual interfaces. You can just assign static-mapped DHCP for each "known" device to put them in a particular piece of the 1 big subnet. Then leave the "unknown" guest devices in the ordinary DHCP pool.
Then your rules can allow/block differently for the different parts of your 1 subnet.
But that provides no real security - any guest can set an IP address themselves, rather than taking DHCP, and effective put themselves in a "more trusted" part of your subnet/rules.
So you need to decide what is your internal known/trusted network, and what are guests and other public stuff.
I suspect that you will want printers and other local network resources (NAS, your own file server…) on the trusted LAN along with your home computers, so they all just see each other.
Many people would end up with:
LAN - your own home computers, printers, NAS, file server, an AP for your home WiFi.
Guest - and AP for your friends to use, with no (or very controlled) access to LAN and more generous internet access
DMZ - anything you have that provides public services (public web site...)
WAN - 1 or more actual uplinks to your ISP/s
So you might end up with 3,4... interfaces on pfSense. If you have that many physical NICs, then easy. Otherwise you need a VLAN switch also.
Then pick some private address space for each of LAN, Guest, DMZ...