Recommendation for I-X/I-Con project DERA-related Demonstration
Austin Tate
Updated: Fri Nov 10 12:08:10 2000
----------------------------------------------------------------------

We agreed to discuss a suitable demo for showing the initial I-X/I-Con
issue-addressing controller design.  The aim is to select a suitable
application at the first I-Con project workshop (1-Dec-2000).

Following discussions between AIAI and DERA, it is suggested that a
new Process Panel is created cleanly based on the I-X design which can
support an "MBP operator" in mapping events to actions they should
take.  It would have a level of process knowledge in a simple process
library, and a way to expand processes which were dependent on the
context or situation that was prevailing at the time. It would be
designed to be one of a variety of different process panels operating
at different time scales and process description levels to support
people involved in military Command and Control in coalitions and
other missions (see notes by Patrick Beautement below on the nature of
different process panels).

The system would perform as follows:

a) Use the issue addressing core of I-Con to handle issues (derived
   from externally generated events) relevant to an MBP user in a
   Coalition C2 process within the context of the CoAX Binni scenario.

b) Where these do not match directly to a known capability, it would
   seek process expansions of how to handle these issues and use a very
   simple expansion engine (a mini-planner really) to match the expanded
   activities to a range of known capabilities which are performable by
   the MBP operator or by other colleagues.  Therefore, in a simple way,
   it would dynamically generate an appropriate response to the issue or
   event in the current situation.

c) It would use the identified actions based on known external
   capabilities to allow the user to "enact" these steps (which in the
   first version would probably correspond to sending messages to
   appropriate external humans or agents).

d) It would maintain a simple display of the current status of issues
   and events delegated to the panel and information on how far along in
   the response process the MBP operator was.

This would bring in the issue addressing engine (without getting
bogged down with too much content in internal specific issue
(internal and external) to be addressed, would look at an internal
process expansion and generative processing facility, and would
address some applications at DERA such as "knowing what to do and who
to contact and inform when an event occurs" (this is something Mark
Round is doing for PJHQ as well).

The suggestion relates to other work that has been discussed with
groups at DERA and could have wider applicability beyond CoAX.

The panel could initially be focused on the MBP/CoAX gathering of
information stage and the I-X/I-Con-based process panel capability in
an early demonstration would, using examples provided by Mike Kirton,
assist a planner in finding an information resource/agent capable of
providing weather in the area of operations; or assist him in what to
do if the weather takes a turn for the worse and the firestorm is not
feasible; assist him in what to do on the discovery of
misinformation, information warfare or denial of service attacks.

-------------------------------------------------------------------

Patrick Beautement makes the following observations on the need for
process panels with different characteristics depending on the user
supported.  He identifies (at least) three types of panel, where the
demonstrator proposed is roughly at his level 2.

1. the JTF HQ's PP is about issues relating to the coherence of the
Joint Plan (time-scale days);

2. the JFAC HQ's PP is about issues relating to the Air Plan (your
MBP operator's - time-scale hours);

3. and there could be a third PP in use by Current Air Operation's
(e.g., Chief of Combat Ops) staff to deal with issues relating to
the execution NOW.

So, in scenario terms: if an Observer observes trucks being where we
didn't expect them to be - this could be of relevance to all three -
but IN DIFFERENT TIMESCALES:

1. Thursday 27 Sep 2012 - JTFC "Oh *@*!*, this Firestorm Option may
have to be excluded";

2. Friday 28 Sep 2012 - JFACC "Oh *@*!*, we've put the Firestorm in
the wrong place - these package / mission options will need updated
target info";

3. Saturday 29 Sep 2012 at 0645 - Chief of Combat Ops "Oh *@*!*,
abort the missions NOW - they're going to the wrong place".