277 lines
24 KiB
Markdown
277 lines
24 KiB
Markdown
---
|
||
tags:
|
||
- '#meetings'
|
||
- '#projects/prin2021'
|
||
alias:
|
||
- "PRIN2021 TODOs"
|
||
---
|
||
|
||
# Notes related to the Prin 2021 proposal
|
||
|
||
- Proposal: [Proposal-PRIN2020 - Google Docs](https://docs.google.com/document/d/1Ygn968RjYRHQ6JQRzdvahrpATZO7pprDl7nBwsOCfNg/edit#)
|
||
- Folder: [Proposta PRIN - Google Drive](https://drive.google.com/drive/u/0/folders/1DyCyvIyNQH_W6IeDC5sob3H80BijiUH9)
|
||
- Titolo del progetto: EMELIOT - Engineered MachinE Learning-intensive IoT systems
|
||
|
||
# Working notes
|
||
- Stiamo puntando ad un TRL 3/4
|
||
- Per quanto riguarda scenari/casi di studio dobbiamo menzionare qualche azienda e che andremo a usare simulatori
|
||
-
|
||
## TODOs
|
||
- Sezione 1 (stato dell'arte)
|
||
- Va messo un cappello iniziale per giustificare le linee di interesse del progetto.
|
||
|
||
- Sezione 2:
|
||
- [x] **Sistemare figura 1 con piramide**
|
||
- [x] CASE STUDY: oltre ad una revisione, richiede una integrazione per quello che riguarda gli esempi di sistemi open source, contatti industriali e simulatori
|
||
|
||
- Sezione 3:
|
||
- [x] rivedere contributi nei task
|
||
- [x] aggiungere una breve descrizione in testa al WP di vostra competenza
|
||
- [x] aggiornare lista expertise e aggiungere una frasetta in cui si spiega il ruolo principale dell’unita’
|
||
- [x] Aggiornare la figura delle dipendenze tra WP
|
||
- [x] Riusciamo a mettere qualche azienda da elencare?
|
||
- Proviamo ad avere un OK per citare qualche azienda?
|
||
- Abbiamo la disponibilità di dati?
|
||
- In WP4 fondere 2,4 e 3,5
|
||
|
||
- Sezione 4
|
||
- [x] Ripassare
|
||
|
||
- [x] O5: da rivedere
|
||
|
||
- GENERALE
|
||
- [x] Controllare gli acronimi
|
||
|
||
|
||
# NOTES
|
||
|
||
## WP DESCRIPTION NOTES
|
||
This WP fulfills O1 i.e., the definition of a low-code development platform to support the design of ML-intensive IoT systems. The different stakeholders of the platform will be provided with domain-specific languages for specifying the components building up the system under development, the data to be managed, and how to exploit them by means of ML techniques. Dedicated recommendation systems will be also defined to support users during such activities. The work is organized around three tasks:
|
||
|
||
## ROLE NOTES
|
||
|
||
- Univaq: model-driven engineering, domain specific languages, low-code development platforms, ML-based recommendation systems, IoT modeling. Main role: Conceive languages and supporting tools for designing IoT systems characterized by ML-based functionalities. In this respect, Univaq will also conceive recommendation systems to help developers to properly configure ML pipelines. Collaborations: Cooperate on the definition of IoT architectures, on the execution of CI/CD pipelines, on test case specifications, and on the development of adversarial ML techniques.
|
||
|
||
## REFINEMENT OF SECTION 4
|
||
[[PRIN2021_SECTION4]]
|
||
|
||
|
||
|
||
|
||
# Call del 19/01/2021
|
||
- [ ] Aggiornare citazioni
|
||
- Sullo stato dell'arte provare a fare un discorso unico
|
||
- [ ] Controllare gli acronimi
|
||
- Durante la call è uscito fuori https://blogs.oracle.com/il-blog-di-oracle/proxima-smart-city3a-la-smart-city-del-futuro-fatta-di-cloud2c-dati-aumentati-e-intelligenza-artificiale-di-oracle
|
||
- [ ] Nella sezione 3 quando parliamo di consorzio menzioniamo attrezzature che abbiamo etc.
|
||
- Riusciamo a mettere qualche azienda da elencare?
|
||
- Proviamo ad avere un OK per citare qualche azienda?
|
||
- Abbiamo la disponibilità di dati?
|
||
|
||
- [ ] Leonardo fa un passaggio sulla 2, noi dobbiamo rivedere la questione case studies
|
||
- [x] **Sistemare figura 1 con piramide**
|
||
- [ ] Rivedere anche Fig WP deps
|
||
- [ ] In WP4 fondere 2,4 e 3,5
|
||
- [ ] Aggiungere contributo nella sezioncina Role and Contribution of the research units
|
||
- [ ] Ripassare sezione 4
|
||
|
||
# Call del 12/01/2021
|
||
|
||
- [ ] Aggiungere chiarimento Fig. 1 per dire che quella e' una architettura di riferimento e che le varie applicazioni istanziano i vari layer con predominanza diversa
|
||
- [ ] Stiamo puntando ad un TRL 3/4
|
||
- [ ] IN OB5 introdurre tra gli obiettivi che andremo a fornire un testbed
|
||
- [ ] Per quanto riguarda scenari/casi di studio dobbiamo menzionare qualche azienda e che andremo a usare simulatori
|
||
- [ ] Raffinare titoli task di WP3
|
||
- [ ] Menzionare il proprio contributo negli altri WP cosi da generare la tabella
|
||
- [ ] Ci sara' poi un paragrafo in cui si dirà in generale il contributo di ogni partner nei vari partner
|
||
- [ ] Nella sezione di impatto fare riferimento agli scenari soprattutti se questi sono importanti da un punt di vista sociale etc. Vedere comunque scheda sulle linee guida per i revisori.
|
||
- [ ] Per quanto riguarda la sezione 1 (stato dell'arte) va messo un cappello iniziale per giustificare le linee di interesse del progetto.
|
||
- [ ] Aggiornare CV e selezionare pubblicazioni.
|
||
|
||
# Call del 5/01/2021
|
||
|
||
- [x] Aggiungere Edge in Fig. 1
|
||
- [x] Guardare [MLOps: Continuous delivery and automation pipelines in machine learning (google.com)](https://cloud.google.com/solutions/machine-learning/mlops-continuous-delivery-and-automation-pipelines-in-machine-learning)
|
||
- [ ] Fare una passata per sottolineare le specificità ML intensive nel contesto IoT rispetto ai sistemi tradizionali.
|
||
- [x] Pensare ad un titolo/acronimo
|
||
- [x] Fare budget [BUDGET_PI_PRIN_2020_c.xlsx - Google Sheets](https://docs.google.com/spreadsheets/d/1jQWdkDvyphpDNmsNn3aL0PlYu0EPTasR/edit#gid=1568281331)
|
||
- 1.2M Euro costo totale del progetto.
|
||
- Non c'e' ne minimo e ne massimo al co-finanziamento.
|
||
- Tabelle stipendiali [[tabelle-definitive-docenti-e-ricercatori-dal-01-01-2019.pdf]]
|
||
- [x] Lavorare su sezione 4
|
||
|
||
# Call del 29/12/2020
|
||
|
||
Far trasparire di piu' la presenza di strumenti a supporto dello sviluppo di sistemi ML-based.
|
||
|
||
- Il nostro principale stackeholder e' l'ingegnere del software che non ha competenze di machine learning.
|
||
- Difficoltà intrinsiche di sistemi IoT
|
||
- Enfatizzare la combinazione di diverse competenze che sono necessarie per sviluppare sistemi IoT che sono intrinsicamente complessi.
|
||
- Necessità di mettere insieme le varie parti insieme
|
||
- Architetture, strumenti model-driven per chiudere il gap tra gli stackholder e gli strumenti necessari
|
||
- Noi ci rivolgiamo ad un team interdisciplinare in cui trovi dentro data scientist, esperto di machine learning, sviluppatore software, etc.
|
||
- Invece di identificare la figura specifica, e' meglio fermarsi a livello di ** team interdisciplinare **
|
||
- Il contesto che sia distribuito sia un data di fatto. E quindi la parte di federated learning va sviluppata a partire da questo. Sempre comunque da un punto di vista ingegneristico (non vogliamo dare contributi nell'aria machine learning). Noi partiamo da modelli machine learning esistenti e vogliamo studiare come poterli integrare in piu' complessi etc. Vediamo l'algoritmo come black box, noi forniamo architetture per comporre, distribuire, etc. Ci sono forse tecniche diverse di learning e quindi potremmo pensare di fornire architetture diverse per supportrarli nei diversi scenari.
|
||
|
||
|
||
## Last notes from Massimiliano
|
||
|
||
### Mail 1
|
||
|
||
```
|
||
Ciao
|
||
|
||
|
||
|
||
per la sezione 3, vi chiederei appena possibile di:
|
||
|
||
\- completare l’individuazione dei task nei WP di vostra competenza. Sono in stato draft e in alcuni WP forse occorre rinominare i task usando nomi + esplicativi
|
||
|
||
\- aggiungere accanto a ogni task degli altri WP se intendete contribuire. Per ora lo si mette cosi’ poi faccio una tabella
|
||
|
||
\- nella sezione "Role and contribution of the research unit” c’e’ un elenco puntato (che poi eventualmente puo’ diventare tabella a seconda dello spazio dove dovremmo provare a elencare le expertise uniche. Mi rendo conto che non e’ banale e che ognuno potrebbe mettere tutto, ma se provassimo a mettere un paio di keyword per ciascuna unita’ (ho messo degli esempi, ma cambiateli/cancellateli) potrebbe funzionare meglio.
|
||
|
||
Se pero’ pensate sia una cattiva idea, lasciamo perdere
|
||
|
||
|
||
|
||
Intanto sto preparando il gantt e una versione aggiornata della figura delle dipendenze tra i wp
|
||
|
||
|
||
|
||
ciao
|
||
|
||
Max
|
||
```
|
||
## Last notes from Leonardo
|
||
|
||
### Mail 1 (Circa sezione 2)
|
||
|
||
Ciao a tutti,
|
||
|
||
ho lavorato ancora alla sezione 2 gestendo i commenti presenti e completandola. Al momento manca la descrizione di un obiettivo e siamo leggermente sopra i 25K caratteri. Dovremo poi ripassarla per portarla a dimensione. Al momento i todo principali sono:
|
||
\- @Luciano: Dovresti scrivere O3 seguendo lo schema indicato (il testo “vecchio” degli obiettivi lo trovi alla fine della sezione 2)
|
||
\- @tutti: come potrete vedere dalla nota che ho aggiunto propongo di togliere la figura su MLops e il mapping degli obiettivi (alla fine quasi tutti gli obiettivi sono ortogonali) tenendo solo il paragrafo del testo che cita la metodologia per dimostrare che siamo aware della sua esistenza e che intendiamo seguirla
|
||
\- @tutti: ho prodotto un draft per i casi study, oltre ad una revisione, richiede una integrazione per quello che riguarda gli esempi di sistemi open source, contatti industriali e simulatori
|
||
\- @chi lavorerà su sicurezza: l’obiettivo O2 è ora focalizzato su testing per reliability (quindi sia aspetti funzionali sia sicurezza). Mi sembra coeso, ma verificate per favore che con le varie revisioni non siano rimaste fuori delle cose importanti (btw che potremmo comunque fare senza scriverle, quindi valutiamo anche l’impatto sulla presentazione del progetto).
|
||
\- @tutti ho scritto un draft di O5 da rivedere
|
||
|
||
Risolti i punti sopra la sezione 2 è pronta. Posso occuparmi io di portarla poi alla lunghezza richiesta.
|
||
|
||
Ciao,
|
||
Leo
|
||
|
||
### Mail 2 (Circa budget)
|
||
|
||
Bozza Budget [[modelloA_2020W3A5FY.pdf]]
|
||
|
||
### Mail 3 (Circa formattazione)
|
||
|
||
Ciao,
|
||
|
||
ho aggiunto i tag html per la formattazione. Proporrei di aggiungerli quando usiamo testo in bold (tra i tag <B> e </B>), in italico (tra i tag <I> e </I>) e bullet list (<ul></ul> perla lista e <li></li> per ciascun item). Li ho già aggiunti in sezione 2. L’obiettivo è poter caricare il testo facendo semplicemente copia e incolla così da tenere sincronizzato il foglio e la versione sul sito, oltre a poter contare i caratteri effettivi.
|
||
|
||
Riguardo ai riferimenti bibliografici, tempo ad un certo potremmo dover optare per una versione molto più compatta: indicare il nome del solo primo autore seguito da et at., indicare la conferenza/rivista per acronimo ed eliminare qualsiasi informazioni non necessaria. A questo punto una correzione da fare all’ultimo una volta ottenuta una stesura abbastanza stabile di tutte le sezioni.
|
||
|
||
Ciao,
|
||
Leo
|
||
|
||
## Struttura della sezione 2
|
||
|
||
- Motivations: _growth of IoT_, gartner emerging challenge for ICT, other paper mentioned by Davide.
|
||
|
||
- _Pervasiveness_ of ML in IoT systems
|
||
|
||
- _Dependability challenge_ and project ambition
|
||
|
||
- Manca un paragrafo che descriva che stiamo parlando di sistemi ML-intensive e che quindi stiamo parlando di un sistema interdisciplinare
|
||
|
||
- [x] Descrizione O1
|
||
|
||
- [x] Sezione 4 sull'impatto
|
||
|
||
## O1 - Development democracy
|
||
|
||
- Methodology: Concerning the _development democracy_ objective, EMELIOT will identify the peculiar elements characterizing ML-Intensive IoT systems. We will design the abstract syntax of a language for modeling such systems and will provide users with constructs to specify for instance data sources, data analysis tasks, machine learning techniques to be employed, etc. The language will be defined through an iterative process consisting of two main steps: a) elicitation of new concepts from the considered application domains, and b) validation of the elicited and properly formalized concepts by modeling concrete AI-Intensive IoT systems.
|
||
- Results: EMELIOT will propose a novel low-code development platform, which will provide adopters with user-friendly graphical interfaces, and drag-and-drop utilities, allowing citizen developers to build ML-Intensive IoT systems. The envisioned platform will be endowed with recommendation systems that will provide teams with recommendations that are relevant for the particular task at hand. For instance, ML experts will get guidance to select and properly configure ML models that might used for the particular IoT system being developed. Similarly, software engineer will get recommendations, e.g., source code snippets, API function calls, etc. which can be employed to develop the IoT system under analysis.
|
||
- Assessment strategy: EMELIOT will assess the benefits of the envisioned low-code development platform within the context of representative use cases. Thus, we aim at involving interested industrial contacts to quantify their experiences in adopting the proposed platform e.g., in terms of development effort, potential cost savings and developer productivity.
|
||
- Relevance of the results for the advancement of knowledge: To the best of our knowledge, there is no concrete methodology and supporting tools for designing and developing ML-intensive IoT systems that particularly focus on fostering the systematic collaboration of interdisciplinary teams consisting of different stakeholders including software engineers, data scientist, and ML experts. Particularly relevant will be the low-code development environment endowed with recommendation systems capable of providing users with building blocks and suggestions, which are relevant for the current development tasks.
|
||
|
||
## Sezione 4 (Impatto): Possible application potentialities and scientific and/or technological and/or social and/or economic impact, with indications of the possible use of research infrastructures
|
||
|
||
EMELIOT will contribute to research and industrial practice by conceiving and assessing innovative solutions for developing, testing, deploying and operating dependable ML-based IoT systems.
|
||
|
||
Internet of Things is rapidly gain ground in different application domains. According to the US National Intelligence Council "by 2025 Internet nodes may reside in everyday things - food packages, furniture, paper documents, and more" [1]. According to a recent report by IoT analytics [2] in 2020, for the first time, IoT connections (i.e., those established by connected cars, smart home devices, industrial equipment, etc.) surpassed non-IoT connections for the first time. According to the same IoT analytics report, by 2025 the expected number of IoT connections will be more than 30 billion, i.e., 4 IoT devices per person on average. With the advent of new like technology standards like 5G, IoT devices are expected to continue to grow much faster.
|
||
|
||
The availability of a massive amount of data from multiple sources is pushing business across the globe to collect and analyze such data. Under such circumstances, according to a recent report [3], Artificial Intelligence and Machine Learning are gaining a key role in IoT enabling the automated identification of patterns, detection of anomalies, and in general they will permit different stakeholders to gain insights from the huge amount of available data. In this respect, the global AI in IoT market size is expected to grow from $ 6.81 billion in 2020 to $18.2 billion in 2025 [3]. Moreover, according to the recent _Global Industry Analysts, Inc_ report [4] the global marked on Machine Learning is expected to reach $27.7 Billion by 2027 growing at a rate of 38.4% over the analysis period 2020-2027. However, according to [3] the lack of skilled workforce, together with concerns related to data security, are some of the major factors, which can restrain the expected market growth.
|
||
|
||
EMELIOT will define techniques and tools to support the whole life-cycle of ML-based IoT systems. The envisioned platform will permit teams consisting of different figures with heterogeneous expertise (e.g., data scientists, ML experts, and software engineers) to collaborate on the development of ML-based systems. The whole software lifecycle will be guided and supported by leveraging MLOps practices. Development environments and recommendation systems will be provided to support the early phases of the system development; advanced testing and monitoring techniques will be devised to check the dependability of the deployed systems. Concerning security aspects, adversarial machine learning techniques will be devised to avoid malfunction in developed machine learning models.
|
||
|
||
Traditionally, data management and analysis are performed by cloud services. However, according to a recent GlobeNewsire report [5], the scenario is changing with the increasing spread of novel technologies, like sensors and other data-producing devices, which permit to perform analysis tasks in the network edge. According to the same report, the edge computing market is expected to grow of 29.4% by 2025. To this end, dedicated techniques and tools are needed to improve the management and analysis of larger datasets in a distributed manner. EMELIOT is expected to provide relevant contributions also in this respect by devising novel architectures underpinning federated and distributed learning techniques. They will permit to optimize ML training activities by mixing the usage of cloud and on-premise services and thus, by properly distributing training tasks on different platforms. The exploitation of edge computing is considered as a key factor to foster the market growth predicted by the _marketdataforecast_ report [3].
|
||
|
||
The scientific and technological impact of the project will be based on top-level publications and demonstrations, and it will be articulated in many directions:
|
||
|
||
(i) A novel low-code environment to model ML-based IoT systems in a systematic way;
|
||
(ii) Newfangled recommendation systems specifically conceived to support different stakeholders, all involved in the development of ML-based IoT systems. Interestingly, the envisioned tools will be able to manage in a homogeneous manner different kinds of artifacts ranging from e.g., data collected from IoT sensors, to snippets of source code that are relevant for the current development task;
|
||
(iii) Innovative approaches and technologies to support the validation and verification of specified ML-based IoT systems;
|
||
(iv) Innovative techniques and tools supporting the continuous integration, delivery and monitoring of ML-intensive IoT systems.
|
||
(v) A novel MLOps infrastructure supporting all the aspects related to the development and operations of ML-intensive IoT systems. The infrastructure will foster a systematic collaboration and interaction of different experts involved in the development of this kind of systems including data scientists, ML experts, and software developers.
|
||
|
||
The EMELIOT partners have identified four communities to foster the use of the envisioned techniques and tools developed in the project:
|
||
|
||
- **Software developers** of IoT systems will create demand for the project technologies as they become aware of the substantial improvements the EMELIOT provides for developing ML-intensive IoT systems. The project will carry out a wide range of dissemination actions to create awareness of the project results amongst Software Developers of IoT systems. As project results are made public, EMELIOT partners expect the dissemination actions will lead to demand for the project technologies amongst Software Developers. This demand will be addressed in some cases by the EMELIOT open source results being adopted directly by some, but for others this demand will be important in motivating commercial providers to adopt and exploit the project results;
|
||
|
||
- **IoT platform providers** are an important part of the EMELIOT exploitation strategy. Once the prototypes are developed and validated within the project, the project partners will target dissemination actions specifically towards IoT platform providers to make them aware of the benefits the project technologies will provide to their customers;
|
||
|
||
- **System Integrators** are a separate community targeted by the EMELIOT as they often deliver IoT solutions for enterprise and government. They often have the capabilities to adopt open source technologies and utilise it within the commercial applications they deliver. The project partners will specifically target System Integrators in dissemination actions and encourage their evaluation and eventual adoption of the EMELIOT technologies for custom application development;
|
||
|
||
- **Academia and Researchers** communities are important because they can contribute to the evolution of the EMELIOT technologies via a sustainable open source process. Dissemination actions will target technical journals, conferences and other academic and research events and communications channels.
|
||
|
||
The partners have the capabilities and existing contacts to address each of the four target markets for the project results.
|
||
|
||
**References**
|
||
|
||
[1] National Intelligence Council, Disruptive Civil Technologies – Six Technologies with Potential Impacts on US Interests Out to 2025 – Conference Report CR 2008-07, April 2008, <http://www.dni.gov/nic/NIC_home.html>
|
||
[2] [State of the IoT 2020: 12 billion IoT connections (iot-analytics.com)](https://iot-analytics.com/state-of-the-iot-2020-12-billion-iot-connections-surpassing-non-iot-for-the-first-time/)
|
||
[3] [Global Artificial Intelligence in IoT Market | Size, Trends, Forecast | 2020 - 2025 (marketdataforecast.com)](https://www.marketdataforecast.com/market-reports/artificial-intelligence-in-iot-market)
|
||
[4] [Machine Learning - Global Market Trajectory & Analytics (researchandmarkets.com)](https://www.researchandmarkets.com/reports/4806169/machine-learning-global-market-trajectory-and?utm_source=dynamic&utm_medium=BW&utm_code=h2m9f6&utm_campaign=1354544+-+Global+Machine+Learning+Market+Analysis%2c+Trends%2c+and+Forecasts+2019-2025&utm_exec=anwr281bwd)
|
||
[5] [Edge Computing Market - Growth, Trends, Forecasts (2020 - 2025) (globenewswire.com)](https://www.globenewswire.com/news-release/2020/12/16/2146204/0/en/Edge-Computing-Market-Growth-Trends-Forecasts-2020-2025.html)
|
||
|
||
## PREVIOUS REFERENCES
|
||
|
||
**
|
||
|
||
### [The power of combining AI and IoT | by Shanika Perera | Towards Data Science](https://towardsdatascience.com/the-power-of-combining-ai-and-iot-4db98ac9f252)
|
||
|
||
[State of the IoT 2018: Number of IoT devices now at 7B – Market accelerating (iot-analytics.com)](https://iot-analytics.com/state-of-the-iot-update-q1-q2-2018-number-of-iot-devices-now-7b/)
|
||
|
||
Spending on machine learning is estimated to reach $57.6 billion by 2021, a compound annual growth rate (CAGR) of 50.1%
|
||
|
||
Minonne, Andrea; Schubmel, David; George, Jebin; Piña, Jeronimo; Danqing Cai, Jessie; Leung, Jonathan; Dimitrov, Lubomir; Ranjan, Manish; Daquila, Marianne; Kumar, Megha; Iwamoto, Naoko; Anand, Nikhil; Carnelley, Philip; Membrila, Roberto; Chaturvedi, Swati; Manabe, Takashi; Vavra, Thomas; Zhang, Xiao-Fei; Zhong, Zhenshan. ["Worldwide Semiannual Artificial Intelligence Systems Spending Guide"](https://www.idc.com/getdoc.jsp?containerId=IDC_P33198). IDC. Retrieved 25 September 2017.
|
||
|
||
### [Gartner 2021 top tech trends](https://drive.google.com/file/d/1Y4xk2Itq6jrEkj6ZU3hi1cRTYa-lIruJ/view?usp=sharing)
|
||
|
||
## **
|
||
|
||
**%Importanza di IoT **
|
||
|
||
> **[State of the IoT 2018: Number of IoT devices now at 7B – Market accelerating (iot-analytics.com)](https://iot-analytics.com/state-of-the-iot-update-q1-q2-2018-number-of-iot-devices-now-7b/)**
|
||
|
||
> **[Gartner 2021 top tech trends](https://drive.google.com/file/d/1Y4xk2Itq6jrEkj6ZU3hi1cRTYa-lIruJ/view?usp=sharing)**
|
||
> **%Importanza di ML e IoT **
|
||
> **[The power of combining AI and IoT | by Shanika Perera | Towards Data Science](https://towardsdatascience.com/the-power-of-combining-ai-and-iot-4db98ac9f252)**
|
||
|
||
> Spending on machine learning is estimated to reach $57.6 billion by 2021, a compound annual growth rate (CAGR) of 50.1%
|
||
|
||
Minonne, Andrea; Schubmel, David; George, Jebin; Piña, Jeronimo; Danqing Cai, Jessie; Leung, Jonathan; Dimitrov, Lubomir; Ranjan, Manish; Daquila, Marianne; Kumar, Megha; Iwamoto, Naoko; Anand, Nikhil; Carnelley, Philip; Membrila, Roberto; Chaturvedi, Swati; Manabe, Takashi; Vavra, Thomas; Zhang, Xiao-Fei; Zhong, Zhenshan. ["Worldwide Semiannual Artificial Intelligence Systems Spending Guide"](https://www.idc.com/getdoc.jsp?containerId=IDC_P33198). IDC. Retrieved 25 September 2017.**
|
||
|
||
**%Importanza di Lowcode nella digitalizzazione**
|
||
|
||
> ...
|
||
|
||
The scientific and technological impact of the project will be based on top-level publications and demonstrations, and it will be articulated in many directions:
|
||
%elencare qui le direzioni relative ai singoli obiettivi
|
||
|
||
**%Altro**
|
||
|
||
> [Amazon, we don’t need another AI tool or APl, we need an open AI platform for cloud and edge | VentureBeat](https://venturebeat.com/2020/12/31/mohammed-farooq-qa-how-hypergiant-is-pushing-for-an-open-ai-platform-for-modelops/amp/) |