Abstract—Ambient Intelligence (AmI, shortly) gathers best back, is capable to modify the corresponding actions
results from three key technologies, Ubiquitous Computing, have been or will be performed.
Ubiquitous Communication, and Intelligent User Friendly Inter-
faces. The functional and spatial distribution of tasks is a natural
We have designed and implemented an intelligent home envi-
thrust to employ multi-agent paradigm to design and implement
AmI environments. Two critical issues, common in most of ap- ronment populated by intelligent appliance agents skilled to
plications, are (1) how to detect in a general and efficient way perform distributed and adaptive “transparent” fuzzy control.
“context” from sensors and (2) how to process contextual infor- The agents interact and coordinate their activities using the
mation in order to improve the functionality of services. In this Fuzzy Mark-up Language [3] (FML) as abstract protocol over
work we experiment a framework where hybrid techniques (dis- shared resources, independently from hardware constraints.
tributed fuzzy control, mobile agents, fuzzy rules induction algo-
The agents compose an abstract layer that binds the instru-
rithms) are mixed to gain flexibility and uniformity.
Keywords: Distributed Fuzzy Control, Mark-up Languages, mental scenario with the services, assuring efficiency and
Agents, Fuzzy Rule Induction Algorithms. adaptivity. This approach allows AmI designers to achieve
useful goals :
I. INTRODUCTION
• to customize the control strategy on specific hardware
When designing AmI Environments [1], different methodolo- through an automatic procedure;
gies and techniques have to be used, ranging from materials • to distribute fuzzy control flow in order to minimize
science, business models, network architectures, up to human global deduction time and better exploit the natural dis-
interaction design. However, as key technologies, AmI is tributed knowledge repositories;
characterizes by: • to acquire, at run time, the user’s behavior and environ-
• Embedded. Devices are (wired or unwired) plugged into ment status in order to apply context-aware adaptivity;
the network. The resulting system consists of several and
multiple devices, compute equipments and software sys- In the remainder of this paper, we will describe our solution
tems that must interact among them. Some of the devices that attempts to solve these complicated problems.
are “simple” sensors, other ones are “actuator” owning a First, we will describe how we have used XML-derived tech-
crunch of control activity on the environment (central- nologies in order to define FML, a markup language skilled
heating, security systems, lightning system, washing ma- for defining detailed structure of fuzzy control independent
chines, refrigerator, etc.). The strong heterogeneity from its legacy representation. Then, we will show the gain
makes difficult a uniformed policy-based management. achievable in terms of distribution and concurrent execution
• Context aware. This term appeared for the first time in by means of agent technology. As last issue, we will describe
[2], where the authors defined context as location, identi- an adaptive methodology, based on inductive computation,
ties of nearby people and objects, and changes to those that allows fuzzy rules to modify themselves according to
objects. Many research groups have been investigating user’s behaviors.
on context-aware applications, but there is no common
understanding what ‘context’ and ‘context awareness’ II. FUZZY MARK-UP LANGUAGE
exactly means. Roughly, the system should own a certain
ability to recognize people and the situational context. Legacy Fuzzy Environment module permits to create a fuzzy
• Personalized. AmI environments are designed for peo- controller using a legacy representation. An example of Leg-
ple, not generic users. This means that the system should acy Fuzzy Environment module is Matlab™ that produce a
be so flexible to tailor itself to meet human needs. .fis file to represent the fuzzy control system. The obtained
• Adaptive. The system, being sensible to the user’s feed- legacy fuzzy controller is passed to the FML Converter mod-
0-7803-9158-6/05/$20.00 © 2005 IEEE. 465 The 2005 IEEE International Conference on Fuzzy Systems
ule that translates it into a markup-based description (fml lan- controller basic concepts of FLC into a tree structure, as
guage). Next step concerns the real implementation of the shown in Figure 1, where each node can be modeled as a
fuzzy controller on specific hardware. XSLT [4] modules are FML tag, and the link father-child represents a nested relation
used to convert fml fuzzy controller in a general purpose between related tags. Currently, we are using FML for model-
computer language using an XSL [5] file containing the trans- ing two well-known fuzzy controllers: Mamdani and Takagi-
lation description. At this level, the control is executable for Sugeno-Kang (TSK) [9].
the hardware.
35
ated to a specific actuator, is executed using the AmI Multi-
Number of Fuzzy Rules
Agents System, as shown in Figure 4. These two phases are
managed by two independent agent classes, so distributed and Centralized Fuzzy Controller Distributed Fuzzy Controller
concurrent processing is achieved, even though some coordi- IV.
nation is required, as better discussed later.
First, it is necessary to defined the feature space on which the Figure 3 - Decentralized vs. Centralized Inference (red
adaptive layer works on. The features, captured in learning line represents the distributed approach, the blue line the
mode, correspond to user’s actions (for example, temperature centralized one)
setting) added with the status of AmI environment. In particu-
lar, each actuator (for instance, a lighting switch) plugged in The adaptive algorithm can be sketched in 7 steps, as shown
the network, communicates its status change in accordance to in figure 6. We refer to the approach described in [10], that
the user’s action and, each sensor (for instance, a lighting sen- we choose as generic model to generate fuzzy rules.