Get Complete Project Material File(s) Now! »

The communicating concrete design

Based on the concept of CM, the ANR McBIM (Material communicating with the BIM – Building Information Model) project aims to design a “communicating concrete” which is equipped with an embedded WSN. The network architecture of communicating concretes is shown in Figure 12. There are two types of nodes: the sensing node (SN) and the communicating node (CN). The SNs measure relevant parameters and transmit measured data to CNs. The CNs then process, store and transfer the received data to the base station.

Position of communicating concrete in the CM paradigm

Concerning the intelligence levels of communicating concrete application in McBIM project, the mentioned classifications in (Meyer et al., 2009) and (Zbib, 2011) could be used to position this application of the CM paradigm, that can be compared with the previous works of (Kubler, 2012) and (Mekki, 2016). Although the behavior of communicating concrete may be different throughout the whole lifecycle, its intelligence on the axis of information processing level, intelligence location and aggregation level stay the same all along its lifecycle. Therefore, the classification of (Meyer et al., 2009) is used to position CM works as depicted in Figure 14.
For the axis level of intelligence, communicating textile can only store product-related information (Information handing level). Thanks to the computing ability of sensor nodes, the communicating concrete developed in (Mekki, 2016) equipped with an embedded WSN can send warning messages to the server (for example, a too high temperature) if the monitoring data exceeds the set value (Problem notification level). As mentioned above, communicating concrete in McBIM project can make decision locally or by their dedicated agent to extend product lifecycle. Therefore, the level of intelligence for CM application in McBIM is the decision-making level.
Mekki proposes network algorithms for data collection, based on routing choices made locally, by each node. Even if it can be locally optimal, the solution is globally suboptimal. In the McBIM Project, the MAS can provide an optimal data collection strategy because it owns a global view of the embedded network. This global strategy could locally be mitigated by the physical node view. This explains why the location of intelligence is between intelligence at object and intelligence through network.
Concerning the aggregation level dimension, composition/decomposition is not considered in the application of communicating textile. For the communicating concrete application in (Mekki, 2016), decomposition has been discussed for data storage and dissemination. In the McBIM project, communicating concrete can work independently by example to report its states during the manufacturing, or group with the others as an aggregate concrete during the construction. As consequence, the aggregation level of communicating concrete in McBIM project is the intelligent container.

Proposed Holonic Approach and related research issues

The digital part example for the management of real materials is shown in Figure 15. Real communicating nodes (CNs) in real concrete are represented by virtual agents in the digital concrete. There are some benefits: firstly, these agents can store monitoring and product-related information to overcome the storage limitation of nodes; Secondly, this digital concrete allows to visualize energy evolution of the embedded WSN based on the received information which can be helpful for the maintaining or the exploitation. Moreover, agents can be grouped into a concrete agent which has the global view of the embedded WSN. This concrete agent can evaluate available collection strategies and then provide energy efficient solutions. By this way, this digital concrete overcomes the limit of processing power at nodes and increases intelligence level of communicating concrete.

Physical part: Energy saving and estimation in WSN

To better understand energy consumption of sensor nodes in WSN, it is necessary to detail the energy consumption of their components. A sensor node has three main components (see Figure 17):
• Sensor: it measures physical phenomena of surrounding environment, for example, temperature, humidity, corrosion, etc. Analogical sensed data are then converted into digital data and transferred to the processor.
• Central Processing Unit (CPU): it is the core part of the sensor node which manages the functionalities of all components (sensors, communication module and so on). Thanks to its memory and information processing capabilities, monitoring data can be stored locally or transmitted to other nodes via the communication unit.
• Communication unit: it has two main functions, receiving and transmitting data from/to other nodes.

Data aggregation energy-efficient routing protocols

A short gathering path helps to reduce the transmission cost from source nodes to BS. As a result, many routing protocols for data collection have been proposed since the last two decades. Concerning the data gathering structure, routing schemes can be classified in two categories: flat routing and hierarchical routing (Khedre et al., 2021). In flat-based routing, messages are flooded between sensors nodes until reaching the sink. In the second one, message transmission follows a hierarchical structure, which allows to define guided paths to the BS and limit the energy consumption of the network. There are three kinds of structures used for hierarchical topology as shown in Figure 23: cluster-based, chain-based and tree-based routing structures.
In a chain-based WSN, nodes transmit information to BS through a linear structure (Lindsey et al., 2002) (Rani et al., 2015). In a cluster-based architecture, few nodes are elected as cluster head (CH) by a selection algorithm for each cluster. Cluster heads aggregate information from other cluster members and transmit aggregated data to BS (Heinzelman et al., 2002) (Braman and Umapathi, 2014). In a tree-based structure, nodes transmit information along a tree structure as in (Tan and Körpeoǧlu, 2005) (Han et al., 2014). Some representative routing protocols from the three structures, adapted to data aggregation, are presented and discussed hereafter.

READ  Directed Graph Degeneracy for Scale-Free Graphs

Digital part: energy management for communicating concrete

To realize the holonic architecture for the communicating concrete, this section begins with the representative traditional holonic control architectures in section 2.2.1 as well as multi-agent architectures for building the informational part of holon in section 2.2.2. Due to the composition characteristic of communicating material and the high flexibility requirement throughout its whole lifecycle, recursive architectures are reviewed in section 2.2.3 to find or design a recursive architecture for the communicating concrete application. At the end, a synthesis is given in section 2.2.4.

Holonic Control Architecture (HCA)

In the manufacturing field, the flexibility of the control system has always been required to adapt to environmental changes. One of the first holonic manufacturing architecture, PROSA (Product-Resource-Oder-Staff) is proposed by Van Brussel (Van Brussel et al., 1998). This architecture consists of three types of basic holons: order holons, product holons, and resource holons. The general architecture is shown in Figure 34. The authors introduced staff holons which can be added to assist the basic holons to complete their work. Holons can cooperate to achieve adaptation and flexibility requirements.

Multi-Agent System (MAS)

In the past two decades, multi-agent system has attracted a lot of attention. Agents can work independently and cooperate to resolve complex tasks. Due to its high flexibility, the agent paradigm has been widely applied in different areas (Sardouk et al., 2013) (Dorri et al., 2018). Congregation (g). The flat is the most basic organization structure, in which all agents can communicate with their neighbors. The agents with common objectives can directly cooperate. In a hierarchical organization, agents are grouped in a tree-like structure, leaf agents can communicate with others via their parent agent (Ma and Zhang, 2014). As the parent agent has knowledge of its children, a local optimal decision can be provided. Meanwhile, the root agent which has a global view could provide optimal solution to face environmental changes. In holonic organization, the agents are considered as the informational part of holons as in (Esmaeili et al., 2017). The holons with the same interest or with same features can be grouped into a higher layer holon. Agents with the same parent can communicate with each other.
In the four other kinds of organization (Team, Coalition, Congregation and Matrix), agents are temporarily or permanently gathered into groups based on their goal. For example, an agent can be part of more than one group in coalition structure (Manisterski et al., 2007), and the different groups communicate with the others via a flat structure to achieve a common goal.

Table of contents :

1.1 Product Lifecycle Management
1.2 Intelligent Product
1.2.1 Definition of intelligent product
1.2.2 Holonic paradigm
1.3 A new type of intelligent product: the communicating material
1.3.1 Definition of the communicating material
1.3.2 Holonic architecture for communicating material
1.3.3 Overview of Wireless Sensor Networks (WSN)
1.3.4 Multi-Agent System
1.4 CM application in the ANR McBIM project
1.4.1 The communicating concrete design
1.4.2 Position of communicating concrete in the CM paradigm
1.5 Proposed Holonic Approach and related research issues
2.1 Physical part: Energy saving and estimation in WSN
2.1.1 In-network data reduction
2.1.2 Data aggregation energy-efficient routing protocols
2.1.3 Time synchronization
2.1.4 Energy consumption models for WSN
2.1.5 Synthesis
2.2 Digital part: energy management for communicating concrete
2.2.1 Holonic Control Architecture (HCA)
2.2.2 Multi-Agent System (MAS)
2.2.3 Recursive holarchy or MAS
2.3 Synthesis
3.1 Energy model of a communicating node
3.2 Data collection in a chain communication structure
3.2.1 Analytical model for data collection without aggregation
3.2.2 Analytical model for Data collection with aggregation
3.3 Analytical model for data aggregation in WSN
3.3.1 Energy consumption estimation method in WSN
3.3.2 Model for tree structure
3.3.3 Model for other structures
3.4 Conclusion
4.1 The proposed CM recursive architecture
4.2 Adapted MAS-R agent model
4.3 Composition / decomposition mechanisms
4.4 Interaction between agents for lifetime estimation
4.5 Cooperation between two aggregated elements
4.6 Energy consumption estimation scenarios
4.6.1 Data collection with lossy data aggregation
4.6.2 Energy consumption estimation for the composition of elements
4.7 Conclusion
5.1 Validation of the energy consumption model for a chain structure
5.1.1 Chain-based data collection platform
5.1.2 Energy consumption of Arduino nodes with XBee shield
5.1.3 Data collection in 3-node chain
5.1.4 Data collection in 6-node chain
5.2 Validation of recursive architecture
5.2.1 Analysis of energy consumption within a single communicating concrete
5.2.2 Using the simulator to analyze the First Node Death time
5.2.3 Cooperation of communicating concretes
5.3 Analysis


Related Posts