Velluscinum: A Middleware for Using Digital Assets in Multi-agent Systems
0 likes62 views
The document presents Velluscinum, a middleware designed for enabling the use of digital assets in multi-agent systems (MAS) with cognitive agents using the BDI model. It highlights the integration of distributed ledger technologies (DLT) with MAS to facilitate agent cooperation, trust management, and direct asset management without needing a certifying authority or external institutions. Furthermore, it details the middleware's capabilities and applications through a case study and emphasizes future research directions involving complex transactions across different MAS.
![• Composed by multiple agents;
These agents can be cognitive, through the use of the
Belief-Desire-Intention (BDI) [4] model;
These agents can interact or compete to resolve a task;
Then, conflicts and uncertainty can occur;
So, it is essential to provide mechanisms for
agents to cooperate by guaranteeing reliability in
the acquiredinformation.
Multi-agent System (MAS)
2
23. Sichman, J.S., Demazeau, Y., Boissier, O.: When can knowledge-based systems be called agents. In: Simpósio Brasileiro de Inteligência Artificial. vol. 9, pp. 172–185.SBC, Rio de Janeiro (1992)
25. Wooldridge, M.J.: Reasoning about rational agents. MIT press (2000)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-2-320.jpg)
![Problem
• Many works integrating DLT +
MAS are still in the theoretical
stage [6].
• We found only one paper [16]
implementing DLT in MAS using
BDI agent.
However, this implementation is in the environmental
dimension of the MAS, making it necessary to create
an institution in the organizational dimension.
7
6. Calvaresi, D., Calbimonte, J.P., Dubovitskaya, A., Mattioli, V., Piguet, J.G., Schumacher, M.: The good, the bad, and the ethical implications of bridging blockchain and multi-agent
systems. Information 10(12) (2019).https://doi.org/10.3390/info10120363
16. Papi, F.G., Hübner, J.F., de Brito, M.: A blockchain integration to support transactions of assets in multi-agent systems. Engineering Applications of Artificial Intelligence 107,104534
(2022).https://doi.org/10.1016/j.engappai.2021.104534](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-7-320.jpg)
![Related Works - Papi et al. (2022)
Integration model between JaCaMo and Ethereum.
Creating a centralized entity and standards allows BDI
agents to request the execution of a smart contract on the
blockchain.
• the implementation is in the environmental dimension
of the MAS;
• it necessary to create an institution in the organizational
dimension (following the Situated Artificial Institution [5]
model)
• the agents can access only one wallet in the DLT,
limiting the competitiveness or autonomy of agents.
• the DLT platform used has high latency and low
performance.
11
16. Papi, F.G., Hübner, J.F., de Brito, M.: A blockchain integration to support transactions of assets in multi-agent systems. Engineering Applications of Artificial Intelligence 107, 104534 (2022).
https://doi.org/10.1016/j.engappai.2021.104534](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-11-320.jpg)
![Theoretical foundation
• A DLT can be considered an append-only
decentralized database because it provides a
storage mechanism.
• However, compared with a traditional database,
its performance is much lower because it has a
low download rate and a high latency [13].
• New approaches based on distributed
databases have been used to improve the
performance of permissioned DLT.
14
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R.,
Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-14-320.jpg)
![Theoretical foundation
• When considering the number of transactions that a DLT
can successfully execute per second:
DLT permissionless (e.g., Bitcoin or Ethereum)
• ranges between 3.45 and 4.69 TPS [24,15];
DLT permissioned (e.g., HyperLedger Fabric)
• vary between 4.28 and 10.51 TPS [7, 15];
DLT permissioned distribution database-based (e.g., BigchainDB)
• varies between 50.60 and 175 TPS [7, 10].
15
7. Contreras, A.F.: Benchmarking of blockchain technologies used in a decentralized data marketplace.
Grado en Ingenierı́a Informática, E.T.S de Ingenieros Informáticos (UPM), Madrid, España (Jun 2019),
https://oa.upm.es/55775/
10. Ge, Z., Loghin, D., Ooi, B.C., Ruan,P., Wang, T.: Hybrid blockchain database systems: Design and
performance. Proc. VLDB Endow. 15(5),1092–1104 (may 2022).
https://doi.org/10.14778/3510397.3510406
15. Nasir, Q., Qasse, I.A., Abu Talib, M., Nassif, A.B.: Performance Analysis of Hyperledger Fabric
Platforms. Security and Communication Networks 2018,1–14 (2018).
https://doi.org/10.1155/2018/3976093
24.Tikhomirov, S.: Ethereum: State of knowledge and research perspectives. In: Imine, A., Fernandez,
J.M., Marion, J.Y., Logrippo, L., Garcia-Alfaro, J. (eds.) Foundations and Practice of Security. pp. 206–221.
Springer International Publishing, Cham (2018).https://doi.org/10.1007/978-3-319-75650-9_14](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-15-320.jpg)
![Theoretical foundation
BigchainDB structures the data as
an owner-controlled asset and
only allows two types of
operation [13]:
• CREATE ASSET
• TRANSFER ASSET
16
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-16-320.jpg)
![Theoretical foundation
A transaction in the BigchainDBare made
up of the following fields [13]:
• ASSET is immutable
information that represents a
register in the DLT;
17
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-17-320.jpg)
![Theoretical foundation
A transaction in the BigchainDB are made
up of the following fields [13]:
• METADATA is additional
information that can be
updated with each new
transaction;
18
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-18-320.jpg)
![Theoretical foundation
A transaction in the BigchainDBare made
up of the following fields [13]:
• INPUT specifies:
which key an asset previously
belonged;
a proof that the conditions
required to transfer ownership
of that asset have been met;
19
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-19-320.jpg)
![Theoretical foundation
A transaction in the BigchainDB are made
up of the following fields [13]:
• OUTPUT specifies the
conditions that need to be
met to change ownership of
a specific asset;
20
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-20-320.jpg)
![Theoretical foundation
A transaction in the BigchainDB are made up of
the following fields [13]:
• Also, a positive number called
AMOUNT is defined in OUTPUT.
• If AMOUNT = 1, the asset is indivisible,
thus representing a non-fungibletoken.
• If AMOUNT > 1, this asset is divisible,
thus representing a token and how many
instances there are;
21
13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white
paper, BigChainDB (2016)](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-21-320.jpg)
![40
https://velluscinum.chon.group/
Extends the Jason-lang [3] through integration with BigchainDB [13],providing
new internal actions to enable the use of digital assets to support the relationship
between BDI agents.
Middleware Velluscinum](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-40-320.jpg)
![Case Study
We adapted the well-known Build-a-House example [1] for
using digital assets in a BigchainDB Network.
• Using the approach of indivisible assets as an agreement.
• Using the approach of divisible assets as cryptocurrencies.
To this, before the creating artifacts phase in the example, we
added plans to deploy an asset for each stage of construction
(representing contracts)
• that are transferreds to the winners before the execution of
the task,
• and are returneds to the owner after the payment
confirmation.
47
Video demonstration: https://youtu.be/qbAU3yD_YhU
1. Boissier, O., Bordini, R.H., Hübner, J.F., Ricci, A., Santi, A.: Multi-agent oriented programming with jacamo.
Science of Computer Programming 78(6),747–761 (2013).https://doi.org/10.1016/j.scico.2011.10.004](https://image.slidesharecdn.com/velluscinum-short-241216182521-225dc37e/85/Velluscinum-A-Middleware-for-Using-Digital-Assets-in-Multi-agent-Systems-47-320.jpg)
Velluscinum: A Middleware for Using Digital Assets in Multi-agent Systems
- 1. Velluscinum: A middleware for using digital assets in Multi-agent Systems Nilson Mori Lazarin1,2, Igor Machado Coelho1, Carlos Eduardo Pantoja1,2 and José Viterbo1 nilson.lazarin@cefet-rj.br 1 Institute of Computing (IC) -- Fluminense Federal University(UFF) 2 Federal Center for TechnologicalEducation Celso Suckow da Fonseca (Cefet/RJ) https://velluscinum.chon.group/
- 2. • Composed by multiple agents; These agents can be cognitive, through the use of the Belief-Desire-Intention (BDI) [4] model; These agents can interact or compete to resolve a task; Then, conflicts and uncertainty can occur; So, it is essential to provide mechanisms for agents to cooperate by guaranteeing reliability in the acquiredinformation. Multi-agent System (MAS) 2 23. Sichman, J.S., Demazeau, Y., Boissier, O.: When can knowledge-based systems be called agents. In: Simpósio Brasileiro de Inteligência Artificial. vol. 9, pp. 172–185.SBC, Rio de Janeiro (1992) 25. Wooldridge, M.J.: Reasoning about rational agents. MIT press (2000)
- 3. Distributed Ledger Technologies (DLT) • Have applications far beyond the financial sector; • Their characteristics: Decentralization; Security; Trust, Low cost of operation. • Then, have a great capacity to contribute in various domains . • They are classified as permissionless or permissioned. 3 9. El Ioini, N., Pahl, C.: A review of distributed ledger technologies. In: On the Move to Meaningful Internet Systems. OTM 2018 Conferences. pp. 277–288.Springer International Publishing, Cham (2018).https://doi.org/10.1007/978-3-030-02671-4_16
- 4. Permissionless DLT • The access is unrestricted; • Members can join and leave the network at any time. 4
- 5. Permissioned DLT • Provides an additional access control layer; • Allowing specific operations by authorized participants. 5
- 6. MAS + DLT • Can facilitate the agreement between agents Taking what is registered in the Ledgeras accurate; • Can be helpful to: Manage trust relationships; Open MAS; Distributed scenarios. 6
- 7. Problem • Many works integrating DLT + MAS are still in the theoretical stage [6]. • We found only one paper [16] implementing DLT in MAS using BDI agent. However, this implementation is in the environmental dimension of the MAS, making it necessary to create an institution in the organizational dimension. 7 6. Calvaresi, D., Calbimonte, J.P., Dubovitskaya, A., Mattioli, V., Piguet, J.G., Schumacher, M.: The good, the bad, and the ethical implications of bridging blockchain and multi-agent systems. Information 10(12) (2019).https://doi.org/10.3390/info10120363 16. Papi, F.G., Hübner, J.F., de Brito, M.: A blockchain integration to support transactions of assets in multi-agent systems. Engineering Applications of Artificial Intelligence 107,104534 (2022).https://doi.org/10.1016/j.engappai.2021.104534
- 8. Mapping review 8 ((“DLT” OR “Distributed Ledger Technologies” OR “Blockchain”) AND (“BDI” OR “belief-desire-intention”) AND (“multi-agent oriented programming” OR “Jason” OR “Jacamo”)) 97 20 *titles and abstracts *involving DLT and MAS *complete reading of the papers 3 *involved DLT and MAS practical applications *search result
- 9. Related Works - Calvaresi et al. (2019) Integration between the MAS and the Hyperledger Fabric. An agent reputation management using smart contract. • Prior registration is required to interact with other agents in the system or operate the ledger, • Use of a third membership service is mandatory. • Agent cannot create an asset in the ledger, only execute a pre-existing contract. • Does not use BDI agents. 9 6. Calvaresi, D., Calbimonte, J.P., Dubovitskaya, A., Mattioli, V., Piguet, J.G., Schumacher, M.: The good, the bad, and the ethical implications of bridging blockchain and multi-agent systems. Information 10(12) (2019).https://doi.org/10.3390/info10120363
- 10. Related Works - Minarsch et al. (2022) A framework for Autonomous Economic Agents (AEA) to operate on Ethereum on behalf of an owner. Objective is to generate economic value for its owner. • Uses an abstraction based on behaviors and handler code. • It does not have BDI-based fundamentals and does not support content based on ontology, agent persistence, and agent mobility services. 10 14. Minarsch,D., Favorito, M., Hosseini, S.A., Turchenkov, Y., Ward, J.: Autonomous economic agent framework. In: Alechina, N., Baldoni, M., Logan, B. (eds.) Engineering Multi-Agent Systems. pp. 237–253.Springer International Publishing, Cham (2022).https://doi.org/10.1007/978-3-030-97457-2_14
- 11. Related Works - Papi et al. (2022) Integration model between JaCaMo and Ethereum. Creating a centralized entity and standards allows BDI agents to request the execution of a smart contract on the blockchain. • the implementation is in the environmental dimension of the MAS; • it necessary to create an institution in the organizational dimension (following the Situated Artificial Institution [5] model) • the agents can access only one wallet in the DLT, limiting the competitiveness or autonomy of agents. • the DLT platform used has high latency and low performance. 11 16. Papi, F.G., Hübner, J.F., de Brito, M.: A blockchain integration to support transactions of assets in multi-agent systems. Engineering Applications of Artificial Intelligence 107, 104534 (2022). https://doi.org/10.1016/j.engappai.2021.104534
- 12. Objective • Using digital assets in the relationships between cognitive agents: the transfer of funds; registration of ownership of artifacts; declarationof promises or agreements. • Allowing the agents to manipulateassets and wallets directly in the DLT. 12
- 13. Related Works X Velluscinum • Unlike all related works we Integrating MAS with BigChainDB; • Unlike Calvaresi et al., this paper eliminates the need for a certifying authority; • Unlike Papi et al., creating a virtual institution is unnecessary, and each agent can manage digital assets directly; • Unlike Minarsch et al., this work allows BDI agents to generate economic value for their owners. 13
- 14. Theoretical foundation • A DLT can be considered an append-only decentralized database because it provides a storage mechanism. • However, compared with a traditional database, its performance is much lower because it has a low download rate and a high latency [13]. • New approaches based on distributed databases have been used to improve the performance of permissioned DLT. 14 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 15. Theoretical foundation • When considering the number of transactions that a DLT can successfully execute per second: DLT permissionless (e.g., Bitcoin or Ethereum) • ranges between 3.45 and 4.69 TPS [24,15]; DLT permissioned (e.g., HyperLedger Fabric) • vary between 4.28 and 10.51 TPS [7, 15]; DLT permissioned distribution database-based (e.g., BigchainDB) • varies between 50.60 and 175 TPS [7, 10]. 15 7. Contreras, A.F.: Benchmarking of blockchain technologies used in a decentralized data marketplace. Grado en Ingenierı́a Informática, E.T.S de Ingenieros Informáticos (UPM), Madrid, España (Jun 2019), https://oa.upm.es/55775/ 10. Ge, Z., Loghin, D., Ooi, B.C., Ruan,P., Wang, T.: Hybrid blockchain database systems: Design and performance. Proc. VLDB Endow. 15(5),1092–1104 (may 2022). https://doi.org/10.14778/3510397.3510406 15. Nasir, Q., Qasse, I.A., Abu Talib, M., Nassif, A.B.: Performance Analysis of Hyperledger Fabric Platforms. Security and Communication Networks 2018,1–14 (2018). https://doi.org/10.1155/2018/3976093 24.Tikhomirov, S.: Ethereum: State of knowledge and research perspectives. In: Imine, A., Fernandez, J.M., Marion, J.Y., Logrippo, L., Garcia-Alfaro, J. (eds.) Foundations and Practice of Security. pp. 206–221. Springer International Publishing, Cham (2018).https://doi.org/10.1007/978-3-319-75650-9_14
- 16. Theoretical foundation BigchainDB structures the data as an owner-controlled asset and only allows two types of operation [13]: • CREATE ASSET • TRANSFER ASSET 16 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 17. Theoretical foundation A transaction in the BigchainDBare made up of the following fields [13]: • ASSET is immutable information that represents a register in the DLT; 17 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 18. Theoretical foundation A transaction in the BigchainDB are made up of the following fields [13]: • METADATA is additional information that can be updated with each new transaction; 18 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 19. Theoretical foundation A transaction in the BigchainDBare made up of the following fields [13]: • INPUT specifies: which key an asset previously belonged; a proof that the conditions required to transfer ownership of that asset have been met; 19 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 20. Theoretical foundation A transaction in the BigchainDB are made up of the following fields [13]: • OUTPUT specifies the conditions that need to be met to change ownership of a specific asset; 20 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 21. Theoretical foundation A transaction in the BigchainDB are made up of the following fields [13]: • Also, a positive number called AMOUNT is defined in OUTPUT. • If AMOUNT = 1, the asset is indivisible, thus representing a non-fungibletoken. • If AMOUNT > 1, this asset is divisible, thus representing a token and how many instances there are; 21 13. McConaghy, T., Marques, R., Müller, A., De Jonghe, D., McConaghy, T., Mc-Mullen, G., Henderson, R., Bellemare, S., Granzotto, A.: Bigchaindb: a scalable blockchain database. white paper, BigChainDB (2016)
- 22. Proposed Approach Use of digital assets to supporting the BDI agents’ relationship: Indivisible assets • as property record • as promise or agreement Divisible assets • as cryptocurrencies 22 Distributed ledger technology MAS - Agents Dimension Velluscinum
- 23. Indivisible asset such as property record 23
- 24. Indivisible asset such as property record 24 • The metadata of asset represents that the artifact can be observed by all agents and used only by agent Bob.
- 25. Indivisible asset such as property record 25 • The asset was transferred to agent Alice, now the artifact could only be observed and used by agent Alice.
- 26. Indivisible asset such as property record 26 • Finally, the asset was transferred to Eve. Now the artifact can be observed by all agents and used only by agent Eve, the current owner of the artifact.
- 27. Indivisible asset such as promise or agreement 27
- 28. Indivisible asset such as promise or agreement 28 • Bob deploy an asset representing a promisse of commitment;
- 29. Indivisible asset such as promise or agreement 29 • Bob transfer the asset to Alice and promisse pay 10 beers;
- 30. Indivisible asset such as promise or agreement 30 • When bob fulfills his commitment and Alice is satisfied, it returns the asset to bob;
- 31. Indivisible asset such as promise or agreement 31 • Later, agent Bob can assume other commitment with Eve.
- 33. Divisible asset such as cryptocurrency 33 • Bob created 100 units of the BobCoin;
- 34. Divisible asset such as cryptocurrency 34 • Bob created 100 units of the BobCoin; • Bob sent 25 units to Alice;
- 35. Divisible asset such as cryptocurrency 35 • Bob created 100 units of the BobCoin; • Bob sent 25 units to Alice; • Bob sent others 25 units to Eve;
- 36. Divisible asset such as cryptocurrency 36 • Alice sent 5 units to Eve;
- 37. Divisible asset such as cryptocurrency 37 • Alice sent 5 units to Eve; • Alice sent others 5 units to Charlie;
- 38. Divisible asset such as cryptocurrency 38 • Eve sent 10 units to Charlie;
- 39. Divisible asset such as cryptocurrency 39 Finally, remained: • Bob has 50; • Alice has 15; • Eve has 20 (5+15); • Charlie has 15 (5+10). Totaling the 100 BobCoins created and distributed by DLT.
- 40. 40 https://velluscinum.chon.group/ Extends the Jason-lang [3] through integration with BigchainDB [13],providing new internal actions to enable the use of digital assets to support the relationship between BDI agents. Middleware Velluscinum
- 41. buildWallet 41
- 42. deployNFT 42
- 43. transferNFT 43
- 44. deployToken 44
- 45. transferToken 45
- 47. Case Study We adapted the well-known Build-a-House example [1] for using digital assets in a BigchainDB Network. • Using the approach of indivisible assets as an agreement. • Using the approach of divisible assets as cryptocurrencies. To this, before the creating artifacts phase in the example, we added plans to deploy an asset for each stage of construction (representing contracts) • that are transferreds to the winners before the execution of the task, • and are returneds to the owner after the payment confirmation. 47 Video demonstration: https://youtu.be/qbAU3yD_YhU 1. Boissier, O., Bordini, R.H., Hübner, J.F., Ricci, A., Santi, A.: Multi-agent oriented programming with jacamo. Science of Computer Programming 78(6),747–761 (2013).https://doi.org/10.1016/j.scico.2011.10.004
- 48. Conclusion • An approach for using digital assets in the relationships between cognitive agents, enabling the representation of: the transfer of funds, registration of property, declaration of promises. • Integration of a well-known example of MAS with DLT was presented using middleware for the agents’ dimension. • Developed and validated a Middleware that provide: each agent can create or transfer assets and sign its transaction through their asymmetric key pair directly in DLT. • Future works can analyze more complex scenarios involving transactions between different multi-agent systems, and the possibilities and implications of a specific permissionless DLT for intelligent agents. 48