Abstract image of a woman sitting on books and studying on a laptop

Digital Twins, Virtual Devices, and Augmentations for Self-Organising Cyber-Physical Collectives

Roberto Casadei, profile picture
Roberto Casadei

The document discusses the concepts of digital twins and virtual nodes within the context of self-organizing cyber-physical collectives, outlining their applications and benefits in decentralized networks. It presents a taxonomy for logical devices, detailing their relationships and execution configurations with physical counterparts. Additionally, it emphasizes the implementation of an architectural model for aggregate computing that facilitates the dynamic integration of virtual devices.

Engineering
Related topics:
Digital Twins
1 of 36
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Digital Twins, Virtual Devices, and Augmentations for Self-Organising Cyber-Physical Collectives [1] Roberto Casadei1 , Danilo Pianini1 , Mirko Viroli1 , Danny Weyns2 1 ALMA MATER STUDIORUM–Università di Bologna, Cesena, Italy 2 Katholieke Universiteit Leuven, Belgium January 13, 2022 Talk @ Fluidware Meeting Campus di Cesena
Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
Digital twins and virtual nodes AC & Gradient-based Navigation Digital twins and virtual devices from literature Digital Twins (DT) in literature Bi-direction synchronisation between digital and physical devices (through so-called digital thread) Virtual Nodes (VN) in literature Work Objectives Techniques Scenarios Topology Type of virtual de- vice [2] Predictability Collaborative emulation of virtual nodes WSN applications Ad-hoc Virtual aggregate [3] Improved QoS TDMA prioritisation WSN applications Clustered Virtual copy [4] Abstraction and efficiency Optimal composition of resource-constrained sensors WSN applications Cloud-based (Hierarchical) Virtual aggregate [5] Improved Qual- ity of Service Quality of Service-aware composition of services Cloud-based IoT applications Cloud-based (Hierarchical) Virtual aggregate [6] Swarm be- haviour control Potential field Mixed swarm- WSN systems Layered Virtual device R.Casadei Background and Motivation Contribution Related work Conclusion References 1/21
Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
Digital twins and virtual nodes AC & Gradient-based Navigation Focus: cyber-physical collectives (CPC) groups of situated agents coordinating to perform some joint/collective task [7] T. De Wolf et al., “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007 R.Casadei Background and Motivation Contribution Related work Conclusion References 2/21
Digital twins and virtual nodes AC & Gradient-based Navigation Focus: cyber-physical collectives (CPC) groups of situated agents coordinating to perform some joint/collective task ∠ using mechanisms like e.g. self-organising information flows [7] [7] T. De Wolf et al., “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007 R.Casadei Background and Motivation Contribution Related work Conclusion References 2/21
Digital twins and virtual nodes AC & Gradient-based Navigation Aggregate Computing (AC) paradigm for CPCs [8] Self-organisation-like programming model for the “artificial collectives” interaction: continuous communication with neighbours behaviour: continuous execution of async rounds of sense – compute – (inter)act abstraction: computational fields [8] M. Viroli et al., “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., 2019 R.Casadei Background and Motivation Contribution Related work Conclusion References 3/21
Digital twins and virtual nodes AC & Gradient-based Navigation Aggregate Computing (AC) paradigm for CPCs [8] Self-organisation-like programming model for the “artificial collectives” interaction: continuous communication with neighbours behaviour: continuous execution of async rounds of sense – compute – (inter)act abstraction: computational fields paradigm: functional — supporting compositionality source destination gradient distance gradient <= + dilate width 37 10 [8] M. Viroli et al., “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., 2019 R.Casadei Background and Motivation Contribution Related work Conclusion References 3/21
Digital twins and virtual nodes AC & Gradient-based Navigation Example: adaptive channel source destination gradient distance gradient <= + dilate width 37 10 few lines of AC code... (leveraging libraries) def channel(src: Boolean, dest: Boolean, width: Double) = dilate(gradient(src) + gradient(dest) <= distance(src, dest), width) def distanceTo(source: Boolean): Double // from lib def distance(source: Boolean, target: Boolean): Double // from lib def dilate(channel: Boolean, width: Double): Boolean // from lib R.Casadei Background and Motivation Contribution Related work Conclusion References 4/21
Digital twins and virtual nodes AC & Gradient-based Navigation Applications and Issues Info flows/channels provide multi-hop connectivity in decentralised, dynamic networks [9] ∠ Applicability: instrastructureless/blackout contexts & dynamic/unknown environments [9] R. Casadei et al., “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 5/21
Digital twins and virtual nodes AC & Gradient-based Navigation Applications and Issues Info flows/channels provide multi-hop connectivity in decentralised, dynamic networks [9] ∠ Applicability: instrastructureless/blackout contexts & dynamic/unknown environments o Assumption: devices as approximation of space Issues: sparseness and (logic) partitions (cf. overcrowded areas) in networks , unreachability – physical or logical (cf. application-level constraints) , long paths → performance/functional issues [9] R. Casadei et al., “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 5/21
Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
Taxonomy for logical devices (1/2) We propose to classify logical devices based on identity and execution relationships with physical devices A taxonomy (by identity correspondence) Correspondence of logical identi- ties with physical identities Logical device (through their soft- ware components) Physical device Description one-to-zero Virtual device – A logical device identifies with no physical de- vice. zero-to-one – Infrastructural device A physical device identifies with no logical de- vice (e.g., it only provides execution support). one-to-one Digital twin Physical twin A logical device identifies with exactly one physical device. one-to-N, N > 1 Virtual aggregate Physical view / copy / member A logical device identifies with a group of physical devices. N-to-one, N > 1 Digital view / copy / member Physical aggregate A physical device identifies with a group of multiple distinct logical devices. R.Casadei Background and Motivation Contribution Related work Conclusion References 6/21
Taxonomy for logical devices (2/2) A taxonomy (by execution relationship) Execution relation- ship of logical de- vices with physical devices Logical device (e.g. via software compo- nents) Physical device Description one-to-zero Undeployed / idle de- vice – The logical device is not executed by any physical device. zero-to-one – Free / thin host The physical device does not (or cannot) host any logical device. one-to-one Embedded logical de- vice Host A logical device is executed by exactly one physical device, namely its host. one-to-N, N > 1 Distributed / Pul- verised device Pulverisation infrastructure element A logical device is executed in a distributed fashion by a group of physical devices. N-to-one, N > 1 Tenant Server A collection of logical devices is executed by an individual physical device. R.Casadei Background and Motivation Contribution Related work Conclusion References 7/21
Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT Aggregate Computing R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT Aggregate Computing Pulverisation Framework R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) ∠ if physical devices are not available, let them be virtual ú augmented collective digital twin δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) ∠ if physical devices are not available, let them be virtual ú augmented collective digital twin ∠ let a managing subsystem handle the self-integration of virtual devices, dynamically δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
PoC Implementation of the Approach (i) we leverage the pulverisation model [10] (an architectural/deployment model for AC) logical device β behaviour χ communication κ state σ sensors α actuators neighbour device χ β κ σ α Host (thin/application-level) Host (thick/infrastructure-level) Logical device σ Device’s sensor set α Device’s actuator set χ Device’s communication interface κ Device’s state β Device’s behaviour Host-to-host link Logical, neighbouring link Twin relationship δ1 δ2 δ3 δ5 δ4 β1 α1 σ1 χ1 κ1 α2 χ2 σ2 β2 κ2 α3 σ3 χ3 β3 κ3 α4 σ4 χ4 β4 κ4 α5 σ5 χ5 β5 κ5 [10] R. Casadei et al., “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, no. 11, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 10/21
PoC Implementation of the Approach (i) we leverage the pulverisation model [10] (an architectural/deployment model for AC) logical device β behaviour χ communication κ state σ sensors α actuators neighbour device χ β κ σ α Host (thin/application-level) Host (thick/infrastructure-level) Logical device σ Device’s sensor set α Device’s actuator set χ Device’s communication interface κ Device’s state β Device’s behaviour Host-to-host link Logical, neighbouring link Twin relationship δ1 δ2 δ3 δ5 δ4 β1 α1 σ1 χ1 κ1 α2 χ2 σ2 β2 κ2 α3 σ3 χ3 β3 κ3 α4 σ4 χ4 β4 κ4 α5 σ5 χ5 β5 κ5 need to virtualise one or more devices (identity, sensor values, neighbourhoods) o this is the challenging part (especially on MANETs) ∠ roughly: build a global map, decide allocations, share virtual state, exploit eventual consistency [10] R. Casadei et al., “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, no. 11, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 10/21
Evaluation (1/6) https://github.com/DanySK/experiment-2021-virtual-devices Figure: Red points = explorers. Yellow square = PoI (target). Blue dots = virtual devices. R.Casadei Background and Motivation Contribution Related work Conclusion References 11/21
Evaluation (2/6) Symbol Meaning Unit Values δj distance error for device j m n.a. (metric) δ mean distance error: P j∈H δj/Nr m n.a. (metric) I isolated devices devices n.a. (metric) Mj comm. cost for device j messages round n.a. (metric) H set of physical devices UIDs adim. n.a. (H ⊂ N) P j∈H Mj total communication cost messages round n.a. (metric) M mean comm. cost: P j∈H Mj/Nr messages round n.a. (metric) Nv virtual device count devices n.a. (metric) Nr physical devices (explorers) devices 5, 15, 50, 158, 500 (Nr = |H|) R logical communication range m 10 · (1, 2, 5, 10, 20, 50, 100, 200) Rv/R relative spawn range adim. 0.25, 0.5, 0.75, 1, 1.25, 1.5, ∞ Rv virtual device spawn range m n.a. (derived as R · Rv/R) Table: Summary of all symbols used in the evaluation R.Casadei Background and Motivation Contribution Related work Conclusion References 12/21
Evaluation (3/6) Plot: mean distance error: δj (distance error; for a physical device j, the difference of distance computed to PoI w.r.t shortest distance from an oracle); R (comm. range); Nr (num of physical devices); Rv /R (relative spawn range); Rv (max distance tolerable before spawning a virtual device in current position—so, ∞ means “no spawning”) 500 1000 1500 2000 2500 3000 3500 time (s) 0 500 1000 1500 2000 2500 Mean distance error (m) Mean distance error when R=50 and Nr=5 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 1000 2000 3000 4000 5000 6000 7000 Mean distance error (m) Mean distance error when R=50 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 1000 2000 3000 4000 5000 Mean distance error (m) Mean distance error when R=100 and Nr=5 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 Mean distance error (m) Mean distance error when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 13/21
Evaluation (4/6) Plot: isolated devices (an isolated device is one for which there is no path to the PoI): R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 10 20 30 40 50 isolated devices (devices) isolated devices when R=10 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 10 20 30 40 50 isolated devices (devices) isolated devices when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 isolated devices (devices) isolated devices when R=10 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 isolated devices (devices) isolated devices when R=100 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 14/21
Evaluation (5/6) Plot: computational cost: Nv (virtual device count—proxy metric for cost); R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 250 500 750 1000 1250 1500 1750 N v (devices) Nv when R=50 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 500 1000 1500 2000 2500 N v (devices) Nv when R=50 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 5 10 15 20 25 30 35 40 N v (devices) Nv when R=1000 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 5 10 15 20 25 30 35 40 N v (devices) Nv when R=1000 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 15/21
Evaluation (6/6) Plot: communication cost: Mj (comm. cost for device j); R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 250 500 750 1000 1250 1500 1750 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=10 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=10 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 600 700 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 14000 16000 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=100 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 16/21
Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
Soft control of collective behaviour by a “shill agent” [11] [11] J. Han et al., “Soft control on collective behavior of a group of autonomous agents by a shill agent,” J. Syst. Sci. Complex., no. 1, 2006 R.Casadei Background and Motivation Contribution Related work Conclusion References 17/21
Swarm behaviour control with WSNs [6] [6] W. Li et al., “Swarm behavior control of mobile multi-robots with wireless sensor networks,” J. Netw. Comput. Appl., no. 4, 2011 R.Casadei Background and Motivation Contribution Related work Conclusion References 18/21
Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
Wrap-up Summary augmented collective digital twins for design and soft control of self-organising cyber-physical collectives Future work devise effective strategies for self-integration of virtual nodes consider generalisations of the idea and their potential, practical impact R.Casadei Background and Motivation Contribution Related work Conclusion References 19/21
Bibliography (1/2) [1] R. Casadei, D. Pianini, M. Viroli, and D. Weyns, “Digital twins, virtual devices, and augmentations for self-organising cyber-physical collectives,” Applied Sciences, vol. 12, no. 1, 2022, ISSN: 2076-3417. DOI: 10.3390/app12010349. [Online]. Available: https://www.mdpi.com/2076-3417/12/1/349. [2] M. Brown, S. Gilbert, N. A. Lynch, C. C. Newport, T. Nolte, and M. Spindel, “The virtual node layer: A programming abstraction for wireless sensor networks,” SIGBED Rev., vol. 4, no. 3, pp. 7–12, 2007. DOI: 10.1145/1317103.1317105. [Online]. Available: https://doi.org/10.1145/1317103.1317105. [3] W. Almobaideen, M. Qatawneh, and O. AbuAlghanam, “Virtual node schedule for supporting qos in wireless sensor network,” in 2019 IEEE Jordan International Joint Conference on Electrical Engineering and Information Technology (JEEIT), IEEE, 2019, pp. 281–285. [4] S. Chatterjee and S. Misra, “Optimal composition of a virtual sensor for efficient virtualization within sensor-cloud,” in International Conference on Communications, IEEE, 2015, pp. 448–453. DOI: 10.1109/ICC.2015.7248362. [Online]. Available: https://doi.org/10.1109/ICC.2015.7248362. [5] M. E. Khansari, S. Sharifian, and S. A. Motamedi, “Virtual sensor as a service: A new multicriteria qos-aware cloud service composition for iot applications,” J. Supercomput., vol. 74, no. 10, pp. 5485–5512, 2018. DOI: 10.1007/s11227-018-2454-y. [Online]. Available: https://doi.org/10.1007/s11227-018-2454-y. R.Casadei Background and Motivation Contribution Related work Conclusion References 20/21
Bibliography (2/2) [6] W. Li and W. Shen, “Swarm behavior control of mobile multi-robots with wireless sensor networks,” J. Netw. Comput. Appl., vol. 34, no. 4, pp. 1398–1407, 2011. DOI: 10.1016/j.jnca.2011.03.023. [Online]. Available: https://doi.org/10.1016/j.jnca.2011.03.023. [7] T. De Wolf and T. Holvoet, “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007, pp. 295–298. [8] M. Viroli, J. Beal, F. Damiani, G. Audrito, R. Casadei, and D. Pianini, “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., vol. 109, p. 100 486, 2019. DOI: 10.1016/j.jlamp.2019.100486. [9] R. Casadei, M. Viroli, G. Audrito, D. Pianini, and F. Damiani, “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020. [10] R. Casadei, D. Pianini, A. Placuzzi, M. Viroli, and D. Weyns, “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, vol. 12, no. 11, 2020. DOI: 10.3390/fi12110203. [Online]. Available: https://doi.org/10.3390/fi12110203. [11] J. Han, M. Li, and L. Guo, “Soft control on collective behavior of a group of autonomous agents by a shill agent,” J. Syst. Sci. Complex., vol. 19, no. 1, pp. 54–62, 2006. DOI: 10.1007/s11424-006-0054-z. [Online]. Available: https://doi.org/10.1007/s11424-006-0054-z. R.Casadei Background and Motivation Contribution Related work Conclusion References 21/21

More Related Content

PDF
Aggregate Computing Platforms: Bridging the Gaps
Roberto Casadei
 
PDF
Bridging the Pervasive Computing Gap: An Aggregate Perspective
Roberto Casadei
 
PDF
A Programming Framework for Collective Adaptive Ecosystems
Roberto Casadei
 
PDF
Scafi: Scala with Computational Fields
Roberto Casadei
 
PDF
Collective Abstractions and Platforms for Large-Scale Self-Adaptive IoT
Roberto Casadei
 
PDF
Tuple-Based Coordination in Large-Scale Situated Systems
Roberto Casadei
 
PDF
Coordinating Computation at the Edge: a Decentralized, Self-organizing, Spati...
Roberto Casadei
 
PDF
Collective Adaptive Systems as Coordination Media: The Case of Tuples in Spac...
Roberto Casadei
 
Aggregate Computing Platforms: Bridging the Gaps
Roberto Casadei
 
Bridging the Pervasive Computing Gap: An Aggregate Perspective
Roberto Casadei
 
A Programming Framework for Collective Adaptive Ecosystems
Roberto Casadei
 
Scafi: Scala with Computational Fields
Roberto Casadei
 
Collective Abstractions and Platforms for Large-Scale Self-Adaptive IoT
Roberto Casadei
 
Tuple-Based Coordination in Large-Scale Situated Systems
Roberto Casadei
 
Coordinating Computation at the Edge: a Decentralized, Self-organizing, Spati...
Roberto Casadei
 
Collective Adaptive Systems as Coordination Media: The Case of Tuples in Spac...
Roberto Casadei
 

What's hot (20)

PDF
On Execution Platforms for Large-Scale Aggregate Computing
Roberto Casadei
 
PDF
From Field-based Coordination to Aggregate Computing
Roberto Casadei
 
PDF
REDUCING RENDER TIME IN RAY TRACING BY PIXEL AVERAGING
ijgca
 
PDF
Reducing Render Time in Ray Tracing by Pixel Averaging
ijcga
 
PPT
DRESD In a Nutshell July07
Marco Santambrogio
 
PDF
FScaFi: A Core Calculus for Collective Adaptive Systems Programming
Roberto Casadei
 
PPTX
Mmsys slideshare-intel-nokia
Rufael Mekuria
 
PDF
ViT (Vision Transformer) Review [CDM]
Dongmin Choi
 
PDF
Indoor Point Cloud Processing
PetteriTeikariPhD
 
PDF
Standardising the compressed representation of neural networks
Förderverein Technische Fakultät
 
PDF
Recognition and Detection of Real-Time Objects Using Unified Network of Faste...
dbpublications
 
PPTX
Fcv rep darrell
zukun
 
PDF
Dataset creation for Deep Learning-based Geometric Computer Vision problems
PetteriTeikariPhD
 
PDF
End-to-End Object Detection with Transformers
Seunghyun Hwang
 
PDF
Interpretability of Convolutional Neural Networks - Eva Mohedano - UPC Barcel...
Universitat Politècnica de Catalunya
 
PPTX
PAKDD2013
Hooman Homayounfard
 
PDF
Cari presentation maurice-tchoupe-joskelngoufo
Mokhtar SELLAMI
 
PDF
SPAR 2015 - Civil Maps Presentation by Sravan Puttagunta
Sravan Puttagunta
 
PDF
TDMA Schleduling in Wireless Sensor Network
neha agarwal
 
PDF
Moldable pipelines for CNNs on heterogeneous edge devices
LEGATO project
 
On Execution Platforms for Large-Scale Aggregate Computing
Roberto Casadei
 
From Field-based Coordination to Aggregate Computing
Roberto Casadei
 
REDUCING RENDER TIME IN RAY TRACING BY PIXEL AVERAGING
ijgca
 
Reducing Render Time in Ray Tracing by Pixel Averaging
ijcga
 
DRESD In a Nutshell July07
Marco Santambrogio
 
FScaFi: A Core Calculus for Collective Adaptive Systems Programming
Roberto Casadei
 
Mmsys slideshare-intel-nokia
Rufael Mekuria
 
ViT (Vision Transformer) Review [CDM]
Dongmin Choi
 
Indoor Point Cloud Processing
PetteriTeikariPhD
 
Standardising the compressed representation of neural networks
Förderverein Technische Fakultät
 
Recognition and Detection of Real-Time Objects Using Unified Network of Faste...
dbpublications
 
Fcv rep darrell
zukun
 
Dataset creation for Deep Learning-based Geometric Computer Vision problems
PetteriTeikariPhD
 
End-to-End Object Detection with Transformers
Seunghyun Hwang
 
Interpretability of Convolutional Neural Networks - Eva Mohedano - UPC Barcel...
Universitat Politècnica de Catalunya
 
PAKDD2013
Hooman Homayounfard
 
Cari presentation maurice-tchoupe-joskelngoufo
Mokhtar SELLAMI
 
SPAR 2015 - Civil Maps Presentation by Sravan Puttagunta
Sravan Puttagunta
 
TDMA Schleduling in Wireless Sensor Network
neha agarwal
 
Moldable pipelines for CNNs on heterogeneous edge devices
LEGATO project
 
Ad

Similar to Digital Twins, Virtual Devices, and Augmentations for Self-Organising Cyber-Physical Collectives (20)

PDF
Augmented Collective Digital Twins for Self-Organising Cyber-Physical Systems
Roberto Casadei
 
PDF
Digital twin
SreelakshmiPerayil
 
PPTX
DIGITAL TWINS 12247944 yasin ali md.pptx
yasinalistudy
 
PDF
Digital_Twin_Technology.pdf
AmrikSingh922379
 
PPTX
Unlocking the Future: Exploring the Power of Digital Twin Technology
22jj5a0511
 
PPTX
Flanders Make Tutorial 2024-Harmony in Complexity_Blended Reasoning in Agent-...
HusseinMarah1
 
PPTX
DIGITAL TWINS - CONCEPT AND APPLICATION -presentation
hrithikanair2003
 
PPTX
How obedient digital twins and intelligent beings contribute to ethics and ex...
Patrick Van Renterghem
 
PPTX
digital twin seminar 1.pptx
MacZain
 
PPTX
Presentation (2) (2).pptx
MohammedSakhlain
 
PDF
The Future Now: Digital Twins in Action
SEOExp
 
PDF
CROM Digital Twins and IoT
Juan C. Vasquez
 
DOCX
1. Digital Twin Technology. thay thien vu
handk22el
 
PDF
Nurturing Digital Twins: How to Build Virtual Instances of Physical Assets to...
Cognizant
 
PDF
Seminar report of digital twin
faheem m m
 
PDF
Digital Twin aiding more effective Digital Maintenance
Diego López-de-Ipiña González-de-Artaza
 
PDF
Digital Twin Technology.pdf
pankaj rajvanshi
 
PPTX
Digital twin - Internet of Things
Ahmed Sayed
 
PPTX
Digital Twin ppt-2 (2).pptx
Vinay Ms
 
PDF
Exploring the Performance and Accuracy of Digital Twin Models
IRJET Journal
 
Augmented Collective Digital Twins for Self-Organising Cyber-Physical Systems
Roberto Casadei
 
Digital twin
SreelakshmiPerayil
 
DIGITAL TWINS 12247944 yasin ali md.pptx
yasinalistudy
 
Digital_Twin_Technology.pdf
AmrikSingh922379
 
Unlocking the Future: Exploring the Power of Digital Twin Technology
22jj5a0511
 
Flanders Make Tutorial 2024-Harmony in Complexity_Blended Reasoning in Agent-...
HusseinMarah1
 
DIGITAL TWINS - CONCEPT AND APPLICATION -presentation
hrithikanair2003
 
How obedient digital twins and intelligent beings contribute to ethics and ex...
Patrick Van Renterghem
 
digital twin seminar 1.pptx
MacZain
 
Presentation (2) (2).pptx
MohammedSakhlain
 
The Future Now: Digital Twins in Action
SEOExp
 
CROM Digital Twins and IoT
Juan C. Vasquez
 
1. Digital Twin Technology. thay thien vu
handk22el
 
Nurturing Digital Twins: How to Build Virtual Instances of Physical Assets to...
Cognizant
 
Seminar report of digital twin
faheem m m
 
Digital Twin aiding more effective Digital Maintenance
Diego López-de-Ipiña González-de-Artaza
 
Digital Twin Technology.pdf
pankaj rajvanshi
 
Digital twin - Internet of Things
Ahmed Sayed
 
Digital Twin ppt-2 (2).pptx
Vinay Ms
 
Exploring the Performance and Accuracy of Digital Twin Models
IRJET Journal
 
Ad

More from Roberto Casadei (20)

PDF
Integrating Collective Computing and the Social Internet of Things for Smart ...
Roberto Casadei
 
PDF
Software Engineering Methods for Artificial Collective Intelligence
Roberto Casadei
 
PDF
Declarative Macro-Programming of Collective Systems with Aggregate Computing:...
Roberto Casadei
 
PDF
Programming (and Learning) Self-Adaptive & Self-Organising Behaviour with Sca...
Roberto Casadei
 
PDF
A Presentation of My Research Activity
Roberto Casadei
 
PDF
Self-Organisation Programming: a Functional Reactive Macro Approach (FRASP) [...
Roberto Casadei
 
PDF
Programming Distributed Collective Processes for Dynamic Ensembles and Collec...
Roberto Casadei
 
PDF
Towards Automated Engineering for Collective Adaptive Systems: Vision and Res...
Roberto Casadei
 
PDF
Aggregate Computing Research: an Overview
Roberto Casadei
 
PDF
Introduction to the 1st DISCOLI workshop on distributed collective intelligence
Roberto Casadei
 
PDF
6th eCAS workshop on Engineering Collective Adaptive Systems
Roberto Casadei
 
PDF
Pulverisation in Cyber-Physical Systems: Engineering the Self-Organising Logi...
Roberto Casadei
 
PDF
Testing: an Introduction and Panorama
Roberto Casadei
 
PDF
On Context-Orientation in Aggregate Programming
Roberto Casadei
 
PDF
Engineering Resilient Collaborative Edge-enabled IoT
Roberto Casadei
 
PDF
Aggregate Processes in Field Calculus
Roberto Casadei
 
PDF
AWS and Serverless Computing
Roberto Casadei
 
PDF
The Rust Programming Language: an Overview
Roberto Casadei
 
PDF
Akka Remoting and Clustering: an Introduction
Roberto Casadei
 
PDF
Akka Actors: an Introduction
Roberto Casadei
 
Integrating Collective Computing and the Social Internet of Things for Smart ...
Roberto Casadei
 
Software Engineering Methods for Artificial Collective Intelligence
Roberto Casadei
 
Declarative Macro-Programming of Collective Systems with Aggregate Computing:...
Roberto Casadei
 
Programming (and Learning) Self-Adaptive & Self-Organising Behaviour with Sca...
Roberto Casadei
 
A Presentation of My Research Activity
Roberto Casadei
 
Self-Organisation Programming: a Functional Reactive Macro Approach (FRASP) [...
Roberto Casadei
 
Programming Distributed Collective Processes for Dynamic Ensembles and Collec...
Roberto Casadei
 
Towards Automated Engineering for Collective Adaptive Systems: Vision and Res...
Roberto Casadei
 
Aggregate Computing Research: an Overview
Roberto Casadei
 
Introduction to the 1st DISCOLI workshop on distributed collective intelligence
Roberto Casadei
 
6th eCAS workshop on Engineering Collective Adaptive Systems
Roberto Casadei
 
Pulverisation in Cyber-Physical Systems: Engineering the Self-Organising Logi...
Roberto Casadei
 
Testing: an Introduction and Panorama
Roberto Casadei
 
On Context-Orientation in Aggregate Programming
Roberto Casadei
 
Engineering Resilient Collaborative Edge-enabled IoT
Roberto Casadei
 
Aggregate Processes in Field Calculus
Roberto Casadei
 
AWS and Serverless Computing
Roberto Casadei
 
The Rust Programming Language: an Overview
Roberto Casadei
 
Akka Remoting and Clustering: an Introduction
Roberto Casadei
 
Akka Actors: an Introduction
Roberto Casadei
 

Recently uploaded (20)

PDF
Abrasive, erosive and cavitation wear.pdf
nfandraden
 
PDF
Accra-Kumasi Expressway - Prefeasibility Report Volume 1 of 7.11.2018.pdf
JeorgeWilsonKingson1
 
PPTX
Chemical Technological Processes, Feasibility Study and Chemical Process Indu...
Raihan87
 
PDF
Artificial Superintelligence (ASI) Alliance Vision Paper.pdf
SonaliPatil325517
 
PDF
BIO-INSPIRED HORMONAL MODULATION AND ADAPTIVE ORCHESTRATION IN S-AI-GPT
ijaia
 
PPTX
tack Data Structure with Array and Linked List Implementation, Push and Pop O...
usham61
 
PDF
737-MAX_SRG.pdf student reference guides
ssusere2119a1
 
PPTX
Module 8- Technological and Communication Skills.pptx
Ramya Nellutla
 
PDF
Human-AI Collaboration: Balancing Agentic AI and Autonomy in Hybrid Systems
ijccsa
 
PPTX
Information Storage and Retrieval Techniques Unit III
jeyamohan2519
 
PPTX
Feature types and data preprocessing steps
deepalishinkar1
 
PDF
Influence of Green Infrastructure on Residents’ Endorsement of the New Ecolog...
Journal of Contemporary Urban Affairs
 
PDF
Soil Improvement Techniques Note - Rabbi
rahmatideal000
 
PDF
Categorization of Factors Affecting Classification Algorithms Selection
IJDKP
 
PDF
UNIT no 1 INTRODUCTION TO DBMS NOTES.pdf
pawarbhaktiit
 
PPTX
Software Engineering and software moduleing
deepikame6
 
PDF
null (2) bgfbg bfgb bfgb fbfg bfbgf b.pdf
Ramez Hosny
 
PDF
August -2025_Top10 Read_Articles_ijait.pdf
ijait
 
PDF
BIO-INSPIRED ARCHITECTURE FOR PARSIMONIOUS CONVERSATIONAL INTELLIGENCE : THE ...
ijaia
 
PPTX
Sorting and Hashing in Data Structures with Algorithms, Techniques, Implement...
usham61
 
Abrasive, erosive and cavitation wear.pdf
nfandraden
 
Accra-Kumasi Expressway - Prefeasibility Report Volume 1 of 7.11.2018.pdf
JeorgeWilsonKingson1
 
Chemical Technological Processes, Feasibility Study and Chemical Process Indu...
Raihan87
 
Artificial Superintelligence (ASI) Alliance Vision Paper.pdf
SonaliPatil325517
 
BIO-INSPIRED HORMONAL MODULATION AND ADAPTIVE ORCHESTRATION IN S-AI-GPT
ijaia
 
tack Data Structure with Array and Linked List Implementation, Push and Pop O...
usham61
 
737-MAX_SRG.pdf student reference guides
ssusere2119a1
 
Module 8- Technological and Communication Skills.pptx
Ramya Nellutla
 
Human-AI Collaboration: Balancing Agentic AI and Autonomy in Hybrid Systems
ijccsa
 
Information Storage and Retrieval Techniques Unit III
jeyamohan2519
 
Feature types and data preprocessing steps
deepalishinkar1
 
Influence of Green Infrastructure on Residents’ Endorsement of the New Ecolog...
Journal of Contemporary Urban Affairs
 
Soil Improvement Techniques Note - Rabbi
rahmatideal000
 
Categorization of Factors Affecting Classification Algorithms Selection
IJDKP
 
UNIT no 1 INTRODUCTION TO DBMS NOTES.pdf
pawarbhaktiit
 
Software Engineering and software moduleing
deepikame6
 
null (2) bgfbg bfgb bfgb fbfg bfbgf b.pdf
Ramez Hosny
 
August -2025_Top10 Read_Articles_ijait.pdf
ijait
 
BIO-INSPIRED ARCHITECTURE FOR PARSIMONIOUS CONVERSATIONAL INTELLIGENCE : THE ...
ijaia
 
Sorting and Hashing in Data Structures with Algorithms, Techniques, Implement...
usham61
 

Digital Twins, Virtual Devices, and Augmentations for Self-Organising Cyber-Physical Collectives

  • 1. Digital Twins, Virtual Devices, and Augmentations for Self-Organising Cyber-Physical Collectives [1] Roberto Casadei1 , Danilo Pianini1 , Mirko Viroli1 , Danny Weyns2 1 ALMA MATER STUDIORUM–Università di Bologna, Cesena, Italy 2 Katholieke Universiteit Leuven, Belgium January 13, 2022 Talk @ Fluidware Meeting Campus di Cesena
  • 2. Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
  • 3. Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
  • 4. Digital twins and virtual nodes AC & Gradient-based Navigation Digital twins and virtual devices from literature Digital Twins (DT) in literature Bi-direction synchronisation between digital and physical devices (through so-called digital thread) Virtual Nodes (VN) in literature Work Objectives Techniques Scenarios Topology Type of virtual de- vice [2] Predictability Collaborative emulation of virtual nodes WSN applications Ad-hoc Virtual aggregate [3] Improved QoS TDMA prioritisation WSN applications Clustered Virtual copy [4] Abstraction and efficiency Optimal composition of resource-constrained sensors WSN applications Cloud-based (Hierarchical) Virtual aggregate [5] Improved Qual- ity of Service Quality of Service-aware composition of services Cloud-based IoT applications Cloud-based (Hierarchical) Virtual aggregate [6] Swarm be- haviour control Potential field Mixed swarm- WSN systems Layered Virtual device R.Casadei Background and Motivation Contribution Related work Conclusion References 1/21
  • 5. Outline 1 Background and Motivation Digital twins and virtual nodes Aggregate Computing & Gradient-based Navigation 2 Contribution 3 Related work 4 Conclusion
  • 6. Digital twins and virtual nodes AC & Gradient-based Navigation Focus: cyber-physical collectives (CPC) groups of situated agents coordinating to perform some joint/collective task [7] T. De Wolf et al., “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007 R.Casadei Background and Motivation Contribution Related work Conclusion References 2/21
  • 7. Digital twins and virtual nodes AC & Gradient-based Navigation Focus: cyber-physical collectives (CPC) groups of situated agents coordinating to perform some joint/collective task ∠ using mechanisms like e.g. self-organising information flows [7] [7] T. De Wolf et al., “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007 R.Casadei Background and Motivation Contribution Related work Conclusion References 2/21
  • 8. Digital twins and virtual nodes AC & Gradient-based Navigation Aggregate Computing (AC) paradigm for CPCs [8] Self-organisation-like programming model for the “artificial collectives” interaction: continuous communication with neighbours behaviour: continuous execution of async rounds of sense – compute – (inter)act abstraction: computational fields [8] M. Viroli et al., “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., 2019 R.Casadei Background and Motivation Contribution Related work Conclusion References 3/21
  • 9. Digital twins and virtual nodes AC & Gradient-based Navigation Aggregate Computing (AC) paradigm for CPCs [8] Self-organisation-like programming model for the “artificial collectives” interaction: continuous communication with neighbours behaviour: continuous execution of async rounds of sense – compute – (inter)act abstraction: computational fields paradigm: functional — supporting compositionality source destination gradient distance gradient <= + dilate width 37 10 [8] M. Viroli et al., “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., 2019 R.Casadei Background and Motivation Contribution Related work Conclusion References 3/21
  • 10. Digital twins and virtual nodes AC & Gradient-based Navigation Example: adaptive channel source destination gradient distance gradient <= + dilate width 37 10 few lines of AC code... (leveraging libraries) def channel(src: Boolean, dest: Boolean, width: Double) = dilate(gradient(src) + gradient(dest) <= distance(src, dest), width) def distanceTo(source: Boolean): Double // from lib def distance(source: Boolean, target: Boolean): Double // from lib def dilate(channel: Boolean, width: Double): Boolean // from lib R.Casadei Background and Motivation Contribution Related work Conclusion References 4/21
  • 11. Digital twins and virtual nodes AC & Gradient-based Navigation Applications and Issues Info flows/channels provide multi-hop connectivity in decentralised, dynamic networks [9] ∠ Applicability: instrastructureless/blackout contexts & dynamic/unknown environments [9] R. Casadei et al., “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 5/21
  • 12. Digital twins and virtual nodes AC & Gradient-based Navigation Applications and Issues Info flows/channels provide multi-hop connectivity in decentralised, dynamic networks [9] ∠ Applicability: instrastructureless/blackout contexts & dynamic/unknown environments o Assumption: devices as approximation of space Issues: sparseness and (logic) partitions (cf. overcrowded areas) in networks , unreachability – physical or logical (cf. application-level constraints) , long paths → performance/functional issues [9] R. Casadei et al., “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 5/21
  • 13. Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
  • 14. Taxonomy for logical devices (1/2) We propose to classify logical devices based on identity and execution relationships with physical devices A taxonomy (by identity correspondence) Correspondence of logical identi- ties with physical identities Logical device (through their soft- ware components) Physical device Description one-to-zero Virtual device – A logical device identifies with no physical de- vice. zero-to-one – Infrastructural device A physical device identifies with no logical de- vice (e.g., it only provides execution support). one-to-one Digital twin Physical twin A logical device identifies with exactly one physical device. one-to-N, N > 1 Virtual aggregate Physical view / copy / member A logical device identifies with a group of physical devices. N-to-one, N > 1 Digital view / copy / member Physical aggregate A physical device identifies with a group of multiple distinct logical devices. R.Casadei Background and Motivation Contribution Related work Conclusion References 6/21
  • 15. Taxonomy for logical devices (2/2) A taxonomy (by execution relationship) Execution relation- ship of logical de- vices with physical devices Logical device (e.g. via software compo- nents) Physical device Description one-to-zero Undeployed / idle de- vice – The logical device is not executed by any physical device. zero-to-one – Free / thin host The physical device does not (or cannot) host any logical device. one-to-one Embedded logical de- vice Host A logical device is executed by exactly one physical device, namely its host. one-to-N, N > 1 Distributed / Pul- verised device Pulverisation infrastructure element A logical device is executed in a distributed fashion by a group of physical devices. N-to-one, N > 1 Tenant Server A collection of logical devices is executed by an individual physical device. R.Casadei Background and Motivation Contribution Related work Conclusion References 7/21
  • 16. Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
  • 17. Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT Aggregate Computing R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
  • 18. Augmented Collective Digital Twins (ACDT) L l0 DT l1 DT l2 DT l3 VN P p0 PT p1 PT p2 PT p3 IN CDT Augmented CDT CPT Augmented CPT Aggregate Computing Pulverisation Framework R.Casadei Background and Motivation Contribution Related work Conclusion References 8/21
  • 19. Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
  • 20. Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) ∠ if physical devices are not available, let them be virtual ú augmented collective digital twin δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
  • 21. Approach (idea) 1) Do not touch the application (aggregate program) 2) Address the issues by changing the system structure (i.e., adding/removing devices) ∠ if physical devices are not available, let them be virtual ú augmented collective digital twin ∠ let a managing subsystem handle the self-integration of virtual devices, dynamically δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx δx R.Casadei Background and Motivation Contribution Related work Conclusion References 9/21
  • 22. PoC Implementation of the Approach (i) we leverage the pulverisation model [10] (an architectural/deployment model for AC) logical device β behaviour χ communication κ state σ sensors α actuators neighbour device χ β κ σ α Host (thin/application-level) Host (thick/infrastructure-level) Logical device σ Device’s sensor set α Device’s actuator set χ Device’s communication interface κ Device’s state β Device’s behaviour Host-to-host link Logical, neighbouring link Twin relationship δ1 δ2 δ3 δ5 δ4 β1 α1 σ1 χ1 κ1 α2 χ2 σ2 β2 κ2 α3 σ3 χ3 β3 κ3 α4 σ4 χ4 β4 κ4 α5 σ5 χ5 β5 κ5 [10] R. Casadei et al., “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, no. 11, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 10/21
  • 23. PoC Implementation of the Approach (i) we leverage the pulverisation model [10] (an architectural/deployment model for AC) logical device β behaviour χ communication κ state σ sensors α actuators neighbour device χ β κ σ α Host (thin/application-level) Host (thick/infrastructure-level) Logical device σ Device’s sensor set α Device’s actuator set χ Device’s communication interface κ Device’s state β Device’s behaviour Host-to-host link Logical, neighbouring link Twin relationship δ1 δ2 δ3 δ5 δ4 β1 α1 σ1 χ1 κ1 α2 χ2 σ2 β2 κ2 α3 σ3 χ3 β3 κ3 α4 σ4 χ4 β4 κ4 α5 σ5 χ5 β5 κ5 need to virtualise one or more devices (identity, sensor values, neighbourhoods) o this is the challenging part (especially on MANETs) ∠ roughly: build a global map, decide allocations, share virtual state, exploit eventual consistency [10] R. Casadei et al., “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, no. 11, 2020 R.Casadei Background and Motivation Contribution Related work Conclusion References 10/21
  • 24. Evaluation (1/6) https://github.com/DanySK/experiment-2021-virtual-devices Figure: Red points = explorers. Yellow square = PoI (target). Blue dots = virtual devices. R.Casadei Background and Motivation Contribution Related work Conclusion References 11/21
  • 25. Evaluation (2/6) Symbol Meaning Unit Values δj distance error for device j m n.a. (metric) δ mean distance error: P j∈H δj/Nr m n.a. (metric) I isolated devices devices n.a. (metric) Mj comm. cost for device j messages round n.a. (metric) H set of physical devices UIDs adim. n.a. (H ⊂ N) P j∈H Mj total communication cost messages round n.a. (metric) M mean comm. cost: P j∈H Mj/Nr messages round n.a. (metric) Nv virtual device count devices n.a. (metric) Nr physical devices (explorers) devices 5, 15, 50, 158, 500 (Nr = |H|) R logical communication range m 10 · (1, 2, 5, 10, 20, 50, 100, 200) Rv/R relative spawn range adim. 0.25, 0.5, 0.75, 1, 1.25, 1.5, ∞ Rv virtual device spawn range m n.a. (derived as R · Rv/R) Table: Summary of all symbols used in the evaluation R.Casadei Background and Motivation Contribution Related work Conclusion References 12/21
  • 26. Evaluation (3/6) Plot: mean distance error: δj (distance error; for a physical device j, the difference of distance computed to PoI w.r.t shortest distance from an oracle); R (comm. range); Nr (num of physical devices); Rv /R (relative spawn range); Rv (max distance tolerable before spawning a virtual device in current position—so, ∞ means “no spawning”) 500 1000 1500 2000 2500 3000 3500 time (s) 0 500 1000 1500 2000 2500 Mean distance error (m) Mean distance error when R=50 and Nr=5 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 1000 2000 3000 4000 5000 6000 7000 Mean distance error (m) Mean distance error when R=50 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 1000 2000 3000 4000 5000 Mean distance error (m) Mean distance error when R=100 and Nr=5 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 Mean distance error (m) Mean distance error when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 13/21
  • 27. Evaluation (4/6) Plot: isolated devices (an isolated device is one for which there is no path to the PoI): R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 10 20 30 40 50 isolated devices (devices) isolated devices when R=10 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 10 20 30 40 50 isolated devices (devices) isolated devices when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 isolated devices (devices) isolated devices when R=10 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 isolated devices (devices) isolated devices when R=100 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 14/21
  • 28. Evaluation (5/6) Plot: computational cost: Nv (virtual device count—proxy metric for cost); R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 250 500 750 1000 1250 1500 1750 N v (devices) Nv when R=50 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 500 1000 1500 2000 2500 N v (devices) Nv when R=50 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 5 10 15 20 25 30 35 40 N v (devices) Nv when R=1000 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 5 10 15 20 25 30 35 40 N v (devices) Nv when R=1000 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 15/21
  • 29. Evaluation (6/6) Plot: communication cost: Mj (comm. cost for device j); R (comm. range); Nr (num of physical devices) 500 1000 1500 2000 2500 3000 3500 time (s) 0 250 500 750 1000 1250 1500 1750 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=10 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=10 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 100 200 300 400 500 600 700 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=100 and Nr=50 Rv R 0.25 0.5 0.75 1 1.25 1.5 500 1000 1500 2000 2500 3000 3500 time (s) 0 2000 4000 6000 8000 10000 12000 14000 16000 Total comm. cost j H M j ( messages round ) Total comm. cost j H Mj when R=100 and Nr=500 Rv R 0.25 0.5 0.75 1 1.25 1.5 R.Casadei Background and Motivation Contribution Related work Conclusion References 16/21
  • 30. Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
  • 31. Soft control of collective behaviour by a “shill agent” [11] [11] J. Han et al., “Soft control on collective behavior of a group of autonomous agents by a shill agent,” J. Syst. Sci. Complex., no. 1, 2006 R.Casadei Background and Motivation Contribution Related work Conclusion References 17/21
  • 32. Swarm behaviour control with WSNs [6] [6] W. Li et al., “Swarm behavior control of mobile multi-robots with wireless sensor networks,” J. Netw. Comput. Appl., no. 4, 2011 R.Casadei Background and Motivation Contribution Related work Conclusion References 18/21
  • 33. Outline 1 Background and Motivation 2 Contribution 3 Related work 4 Conclusion
  • 34. Wrap-up Summary augmented collective digital twins for design and soft control of self-organising cyber-physical collectives Future work devise effective strategies for self-integration of virtual nodes consider generalisations of the idea and their potential, practical impact R.Casadei Background and Motivation Contribution Related work Conclusion References 19/21
  • 35. Bibliography (1/2) [1] R. Casadei, D. Pianini, M. Viroli, and D. Weyns, “Digital twins, virtual devices, and augmentations for self-organising cyber-physical collectives,” Applied Sciences, vol. 12, no. 1, 2022, ISSN: 2076-3417. DOI: 10.3390/app12010349. [Online]. Available: https://www.mdpi.com/2076-3417/12/1/349. [2] M. Brown, S. Gilbert, N. A. Lynch, C. C. Newport, T. Nolte, and M. Spindel, “The virtual node layer: A programming abstraction for wireless sensor networks,” SIGBED Rev., vol. 4, no. 3, pp. 7–12, 2007. DOI: 10.1145/1317103.1317105. [Online]. Available: https://doi.org/10.1145/1317103.1317105. [3] W. Almobaideen, M. Qatawneh, and O. AbuAlghanam, “Virtual node schedule for supporting qos in wireless sensor network,” in 2019 IEEE Jordan International Joint Conference on Electrical Engineering and Information Technology (JEEIT), IEEE, 2019, pp. 281–285. [4] S. Chatterjee and S. Misra, “Optimal composition of a virtual sensor for efficient virtualization within sensor-cloud,” in International Conference on Communications, IEEE, 2015, pp. 448–453. DOI: 10.1109/ICC.2015.7248362. [Online]. Available: https://doi.org/10.1109/ICC.2015.7248362. [5] M. E. Khansari, S. Sharifian, and S. A. Motamedi, “Virtual sensor as a service: A new multicriteria qos-aware cloud service composition for iot applications,” J. Supercomput., vol. 74, no. 10, pp. 5485–5512, 2018. DOI: 10.1007/s11227-018-2454-y. [Online]. Available: https://doi.org/10.1007/s11227-018-2454-y. R.Casadei Background and Motivation Contribution Related work Conclusion References 20/21
  • 36. Bibliography (2/2) [6] W. Li and W. Shen, “Swarm behavior control of mobile multi-robots with wireless sensor networks,” J. Netw. Comput. Appl., vol. 34, no. 4, pp. 1398–1407, 2011. DOI: 10.1016/j.jnca.2011.03.023. [Online]. Available: https://doi.org/10.1016/j.jnca.2011.03.023. [7] T. De Wolf and T. Holvoet, “Designing self-organising emergent systems based on information flows and feedback-loops,” in First International Conference on Self-Adaptive and Self-Organizing Systems (SASO 2007), IEEE, 2007, pp. 295–298. [8] M. Viroli, J. Beal, F. Damiani, G. Audrito, R. Casadei, and D. Pianini, “From distributed coordination to field calculus and aggregate computing,” J. Log. Algebraic Methods Program., vol. 109, p. 100 486, 2019. DOI: 10.1016/j.jlamp.2019.100486. [9] R. Casadei, M. Viroli, G. Audrito, D. Pianini, and F. Damiani, “Engineering collective intelligence at the edge with aggregate processes,” Engineering Applications of Artificial Intelligence, 2020. [10] R. Casadei, D. Pianini, A. Placuzzi, M. Viroli, and D. Weyns, “Pulverization in cyber-physical systems: Engineering the self-organizing logic separated from deployment,” Future Internet, vol. 12, no. 11, 2020. DOI: 10.3390/fi12110203. [Online]. Available: https://doi.org/10.3390/fi12110203. [11] J. Han, M. Li, and L. Guo, “Soft control on collective behavior of a group of autonomous agents by a shill agent,” J. Syst. Sci. Complex., vol. 19, no. 1, pp. 54–62, 2006. DOI: 10.1007/s11424-006-0054-z. [Online]. Available: https://doi.org/10.1007/s11424-006-0054-z. R.Casadei Background and Motivation Contribution Related work Conclusion References 21/21