everything to be said is already said..
it is an industrial type of network. it allows data exchange between nodes. you can think of it as ProfiBUs over Ethernet. like all industrial networks, to optimize efficiency special protocol is used but hardware and network infrastructure for industrial ethernet network is no different than plain Ethernet you may use at home or in the office.
all networks support standard data ("standard" means not safety rated). transferred message is also called frame or a telegram or mail or packet etc. but they all mean the same thing. basically message contains two things:
a) some overhead needed for transport (info like, message number, sent from, sent to, size etc.).
b) data-block or "payload".. this is actual DATA being transmitted. it will be extracted when transferred message reaches destination
some networks also support safety frame. normally this is a separate message (with different structure) that is sent at different interval since high priority. failure to arrive on time (as measured from previous exchange) would be considered loss, if certain number of messages are lost, connection is deemed bad and safety node shuts down assuming the worst... number of consecutive messages that trigger fault are some small number, usually 1,2,3 or so.
in standard Ethernet networks nodes are called client or server based on their role in the network.
server has to start first, then it sits and waits for some request. when request is made, server tries to process it and return result, then waits for next request. any website is an example of a server. client is device that tries to connect to server. web browser of your computer is one such client.
but anyone has seen some pages loading slowly when server is too busy or network is exhausted, but... the system is able to deliver messages though a variety of routes, if reconnected, new route will need to be reconnected. such features add complexity and resolve problems even after some deays or several retries.
industrial networks use same hardware but protocols are designed to not be as flexile (many things are stripped down) but optimized for performance (fast reconnect, fast update, little or no negotiation etc.).
also terminology is changed, instead of client/server, there are terms like master/slave or controller/device or scanner/adapter etc.
role is important because it defines resources device may need.
slave is usually a "stupid" device with fixed purpose... like IO
.... slaves are normally not aware of other slaves and don't care about them .. .as long as everyone has unique ID (name or address or node number)
slaves are a lot like servers. they sit and wait until someone tells them to do something.
master is different. it needs to know how many slaves are in his/its control. what are their capabilities (IO type and size) etc. master is one that starts all communication (establish connection, tell slaves what to do etc.).
so to make an industrial network you need a master... and one or more slaves.
in general master-to-master communication does not work. same is with slave-to-slave.
so your robots were part of a plant that used Siemens Safety PLC. Safety PLC can process both standard and safety logic and in same network can be mix of slaves, some that only support standard messages, some that support only safety messages, some that support both.
that PLC is apparently still in the plant and not with you. it was working as a master for both standard messages and for safety messages.
your robots were slaves. from PLC side they looked like dumb devices (nothing fancy, simple blocks of inputs and outputs). actually there are two sets of IO blocks (standard and safety).
now that robots are standalone, they don't get messages from master so you get ProfiNet messages complaining that master is not contacting the robot. worse, safety messages are missing. this means that safety circuit in robot times out and then disables drives. this is why you cannot power up robot drives.
unless you connect (and properly configure) safety PLC to act as a master, your robots will be dead. , same master can be connected to several robots (one master can connect to several slaves). but you are also lucky because with KRC4 you can still at least play with robots in StartUp mode (requires T1 mode) but you will NOT be able to use robots in AUT or EXT mode.
so you are looking at getting safety rated PLC, Siemens software, some safety IO and having a lot of fun learning how to use it. But this is a good news... it has lots of potential.
the other news is that KRC4s can have different safety interfaces. one of them is X11. when using X11, ProfiNET does not need to be uninstalled... but it need to be at least configured differently. for example you can use it to exchange only standard data with Siemens PLC, instead of safety messages (ProfiSafe) you could use X11. or you could completely get rid of ProfiNet for now but you must have functional safety interface (this is used to enable drives).
if you do not have X11 interface uninstalling ProfiNet will disable robot completely... even Startup Mode will no longer work (instead of red banner, you will see gray one). gray banner means no valid safety interface is found so drives cannot be enabled. StartupMode cannot enable drives on its own... it only bypasses safety interface. but ... when no interface is configured - there is nothing to bypass.
using X11 interface means hardwiring safety devices (external inputs, Operator safety, Operator safety Ack button, external enabling switch). For test one can place some jumpers. This is dangerous of course but at least this would allow enabling of drives any time and running in any mode - without safety PLC.