DPRG
DPRG List  



[DPRG] Is subsumption old hat?

Subject: [DPRG] Is subsumption old hat?
From: Randy M. Dumse rmd at newmicros.com
Date: Mon Sep 10 09:39:19 CDT 2007

William Crolley wrote: Saturday, September 08, 2007 11:17 AM
> The new sensor architecture that is competing with 
> subsumption would be the various types of sensor fusion. 
> ... This is a major problem for subsumption which works 
> best with binary info, similar to relay logic.

Very insightful comment, I thought. Certaily the trigger of the
behavior in Subsumption is rather binary. That is, the behavior
is triggered or is not, often based on a preset level.

However even Jones in his book allows for some preprocessing of
signals. So once again, we might be up against the vaggeries of
Subsumption. Is preprocesssing allowed? I didn't see any such
evidence in Brooks writings. So is Jones version a true form? Or
a variant? Or would Brooks see the answer as self evidently a
detail not affecting or connecting to the Subsumption ideas at
all.

I've even wondered about Jones "Leaky Integrator" code in his
Appendicies. This is a hidden form of including state
information (as evidenced by the integrator depending on the
history of previous inputs). To me this seems a very deliberte
departure from Brooks.

Randy
www.newmicros.com


More information about the DPRG mailing list