Papers by Andrea de Giorgio
Journal of Manufacturing Systems
Sensors
Surgical simulation practices have witnessed a rapid expansion as an invaluable approach to resid... more Surgical simulation practices have witnessed a rapid expansion as an invaluable approach to resident training in recent years. One emerging way of implementing simulation is the adoption of extended reality (XR) technologies, which enable trainees to hone their skills by allowing interaction with virtual 3D objects placed in either real-world imagery or virtual environments. The goal of the present systematic review is to survey and broach the topic of XR in neurosurgery, with a focus on education. Five databases were investigated, leading to the inclusion of 31 studies after a thorough reviewing process. Focusing on user performance (UP) and user experience (UX), the body of evidence provided by these 31 studies showed that this technology has, in fact, the potential of enhancing neurosurgical education through the use of a wide array of both objective and subjective metrics. Recent research on the topic has so far produced solid results, particularly showing improvements in young ...
Questo articolo scientifico sostiene la tesi che un algoritmo di controllo, facente uso di intell... more Questo articolo scientifico sostiene la tesi che un algoritmo di controllo, facente uso di intelligenza artificiale, per poter essere efficiente necessiti di sfruttare le simmetrie integrate in ogni manipolatore robotico industrale così che quest'ultimo possa essere ulteriormente caratterizzato ed utilizzato. Il prodotto di questo miglioramento è uno spazio discreto a griglia cilindrica quadridimensionale (4D) che può direttamente sostituire complessi modelli robotici. A* è scelto per il suo ampio utilizzo tra simili algoritmi di ricerca, in modo da studiare i vantaggi e gli svantaggi del controllo di robot industriali tramite la griglia discreta cilindrica 4D. Lo studio mostra che questo approccio consente di controllare un robot senza alcuna conoscenza specifica dei modelli cinematici e dinamici del robot al momento della pianificazione e dell'esecuzione. In effetti, le posizioni dei giunti del robot per ciascuna cella della griglia vengono precalcolate e memorizzate come ...
Industrial processes are mainly based on procedural knowledge that must be continually elicited f... more Industrial processes are mainly based on procedural knowledge that must be continually elicited from experienced operators and learned by novice operators. In the context of Industry 4.0, machines already play a key role in knowledge transfer; however, new models and methods based on the artificial intelligence advances of the past few years need to be developed and applied. The future of human-machine collaboration is not limited to physical applications, but it has the potential to harness both the strength of human skills, experience and the computational power provided by the surrounding machines for truly adaptive industrial processes. The winning recipe is a balance between letting humans exploit their inherent experience and letting machines integrate the missing skills to preserve production standards. This work introduces a procedural knowledge model to be used for the design of industrial and scientific adaptive processes and it paves the way to transforming human-machine collaboration into an efficient solution to make industrial and scientific processes resilient to a constantly changing world.Industriella processer baseras huvudsakligen på den procedurella kunskapen som fortlöpande måste tas fram och anpassas av erfarna operatörer och läras in av nybörjare. Inom ramen för Industri 4.0 spelar maskiner redan en nyckelroll i kunskapsöverföring; dock behöver nya modeller och metoder utvecklas och användas, som baseras på de senaste årens framsteg inom artificiell intelligens. Framtiden för samarbete mellan människa och maskin är inte begränsad till fysiska applikationer, utan den har potential att utnyttja såväl styrkan i mänsklig kompetens och erfarenhet som den beräkningskraft som de omgivande maskinerna tillhandahåller, för att åstadkomma verkligt anpassningsbara industriella processer. Det vinnande receptet är att hitta en balans mellan att låta människor utnyttja sina egna erfarenheter och att låta maskiner tillhandahålla de saknade färdigheterna för att kunna följa produktionsstandarder. I detta arbete introduceras en procedurell kunskapsmodell som kan användas för utformning av industriella och vetenskapliga, anpassningsbara processer och banar väg för att omvandla samarbete mellan människor och maskiner till effektiva lösningar för att göra industriella och vetenskapliga processer följsamma i en ständigt föränderlig värld
This article is a citation review of the publication "A hybridization of genetic algorithms ... more This article is a citation review of the publication "A hybridization of genetic algorithms and fuzzy logic for the single-machine scheduling with flexible maintenance problem under human resource ...
This article argues that despite a citation review is a rarely used research tool, this can be ve... more This article argues that despite a citation review is a rarely used research tool, this can be very useful to assess the impact of new research topics, both from the future research direction and the bibliometric perspectives. An explorative study is presented around the research area marked as Industry 4.0 with the conference paper mentioned in the title of this citation review. Even though the given reference paper is relatively recent, there are already twenty-seven citations listed among three different scholar databases. These are Google Scholar, ResearchGate and Semantic Scholar. In light of this, the article provides a bibliometric confirmation and analysis for the progression of the line of research adopted by de Giorgio et al. in the exploration of non-traditional methods using virtual reality technology and human-robot collaboration for adaptive applications in Industry 4.0. Furthermore, it represents a model for the authors’ self-development and an example of an unconvent...
Procedia Manufacturing, 2020
In the age of digital manufacturing there is a need to elicit and transfer procedural knowledge b... more In the age of digital manufacturing there is a need to elicit and transfer procedural knowledge between humans and machines. Having proper knowledge is essential in decision-making. The more the knowledge, the better decisions are made. To capture experiences and turn them into knowledge is fundamental in learning processes and knowledge development. Knowledge engineering and knowledge management have been subject for research for decades and several concepts about knowledge and knowledge transfer have been introduced, but a functional approach to exploit knowledge efficiently in manufacturing is still missing. In the era of Industry 4.0, humans and machines must be able to collaborate in such way that both can exploit the best abilities of each other in a manufacturing process. This paper introduces a procedural knowledge process (PKP) approach to capturing and defining unexpected events, while a process step is able to perform its required functions and transfer that information as machine-understandable knowledge about a failure mode. Function blocks (FBs), as per the IEC-61499 standard, have been proposed as a way to achieve distributed process planning in which the manufacturing process can adapt itself to runtime conditions, e.g. machine availability, etc. However, FBs are event-driven systems and the approach is limited to work under well-known runtime conditions, e.g. machine configurations and states, or deviations which are impossible to foresee in advance, for instance the outcome of a process failure mode effects analysis (PFMEA). The PKP introduced in this paper, aims at bridging this gap by integrating at runtime an expert operator's solution based on root cause analysis (RCA) in an FB architecture, making the FB knowledge-driven systems, for further executions of the same without redesigning it. Natural language representations of procedural knowledge blocks (PKBs) allow to transfer procedural knowledge to human operators, i.e. explain the process flow of a machine decision, while machine representations of PKBs allow to embed procedural knowledge that is elicited from expert operators upon unexpected events into the FBs process. The resulting PKP enhances the FBs for smart industrial applications, such as the process planning use case described in this paper.
Journal of Manufacturing Systems, 2021
Literature shows that reinforcement learning (RL) and the well-known optimization algorithms deri... more Literature shows that reinforcement learning (RL) and the well-known optimization algorithms derived from it have been applied to assembly sequence planning (ASP); however, the way this is done, as an offline process, ends up generating optimization methods that are not exploiting the full potential of RL. Today's assembly lines need to be adaptive to changes, resilient to errors and attentive to the operators' skills and needs. If all of these aspects need to evolve towards a new paradigm, called Industry 4.0, the way RL is applied to ASP needs to change as well: the RL phase has to be part of the assembly execution phase and be optimized with time and several repetitions of the process. This article presents an agile exploratory experiment in ASP to prove the effectiveness of RL techniques to execute ASP as an adaptive, online and experience-driven optimization process, directly at assembly time. The human-assembly interaction is modelled through the input-outputs of an assembly guidance system built as an assembly digital twin. Experimental assemblies are executed without pre-established assembly sequence plans and adapted to the operators' needs. The experiments show that precedence and transition matrices for an assembly can be generated from the statistical knowledge of several different assembly executions. When the frequency of a given subassembly reinforces its importance, statistical results obtained from the experiments prove that online RL applications are not only possible but also effective for learning, teaching, executing and improving assembly tasks at the same time. This article paves the way towards the application of online RL algorithms to ASP.
Procedia Manufacturing, 2017
This paper outlines the main steps towards an open and adaptive simulation method for human-robot... more This paper outlines the main steps towards an open and adaptive simulation method for human-robot collaboration (HRC) in production engineering supported by virtual reality (VR). The work is based on the latest software developments in the gaming industry, in addition to the already commercially available hardware that is robust and reliable. This allows to overcome VR limitations of the industrial software provided by manufacturing machine producers and it is based on an open-source community programming approach and also leads to significant advantages such as interfacing with the latest developed hardware for realistic user experience in immersive VR, as well as the possibility to share adaptive algorithms. A practical implementation in Unity is provided as a functional prototype for feasibility tests. However, at the time of this paper, no controlled human-subject studies on the implementation have been noted, in fact, this is solely provided to show preliminary proof of concept. Future work will formally address the questions that are raised in this first run.
Adaptive Behavior, 2015
Distributed and hierarchical models of control are nowadays popular in computational modeling and... more Distributed and hierarchical models of control are nowadays popular in computational modeling and robotics. In the artificial neural network literature, complex behaviors can be produced by composing elementary building blocks or motor primitives, possibly organized in a layered structure. However, it is still unknown how the brain learns and encodes multiple motor primitives, and how it rapidly reassembles, sequences and switches them by exerting cognitive control. In this paper we advance a novel proposal, a hierarchical programmable neural network architecture, based on the notion of programmability and an interpreter-programmer computational scheme. In this approach, complex (and novel) behaviors can be acquired by embedding multiple modules (motor primitives) in a single, multi-purpose neural network. This is supported by recent theories of brain functioning in which skilled behaviors can be generated by combining functional different primitives embedded in “reusable” areas of ...
This article argues that an efficient artificial intelligence control algorithm needs the built-i... more This article argues that an efficient artificial intelligence control algorithm needs the built-in symmetries of an industrial robot manipulator to be further characterized and exploited. The product of this enhancement is a four-dimensional (4D) discrete cylindrical grid space that can directly replace complex robot models. A* is chosen for its wide use among such algorithms to study the advantages and disadvantages of steering the robot manipulator within the 4D cylindrical discrete grid. The study shows that this approach makes it possible to control a robot without any specific knowledge of the robot kinematic and dynamic models at planning and execution time. In fact, the robot joint positions for each grid cell are pre-calculated and stored as knowledge, then quickly retrieved by the pathfinding algorithm when needed. The 4D cylindrical discrete space has both the advantages of the configuration space and the three-dimensional Cartesian workspace of the robot. Since path optim...
This is the version of the code released with the scientific article A. de Giorgio and L. Wang, &... more This is the version of the code released with the scientific article A. de Giorgio and L. Wang, "Artificial Intelligence Control in 4D Cylindrical Space for Industrial Robotic Applications", in IEEE Access, vol. 8, pp. 174833-174844, 2020, doi: 10.1109/ACCESS.2020.3026193.
Obiettivo della mia tesi è studiare come la presenza di circuiti neurali programmabili possa rend... more Obiettivo della mia tesi è studiare come la presenza di circuiti neurali programmabili possa rendere l'apprendimento di differenti comportamenti, da parte di una singola struttura, più efficace. A tal fine, ho messo a confronto due architetture, una programmabile e l'altra non programmabile, come parte di uno scenario robotico appositamente progettato per testare la possibilità di apprendere ed esibire comportamenti multipli. Più in particolare, ho supposto che un robot possa imparare, tramite la sua architettura di controllo, programmabile o meno, a raggiungere, su richiesta, ciascuna delle otto terminazioni differenti di un labirinto costituito da una serie di tre biforcazioni e di dimensioni variabili. Fondamentale, in tale scenario, la necessità per l'architettura di controllo di apprendere, e poi esibire, comportamenti e non semplici traiettorie. In tal modo, ho potuto raccogliere dati sufficienti per realizzare un'opportuna analisi.
In this thesis I demonstrated how a singular neural network can potentially represent the set of ... more In this thesis I demonstrated how a singular neural network can potentially represent the set of more latent neural circuits, able to execute different functions, based on an input encoding so to reprogram their functionality. Such programmable structure, in passing from one behavior to another, does not require further learning procedures, nor any structural modifications. This is ideal to execute quick and reversible transitions such as the ones showed in biological organisms’ behaviors.
Restricted Boltzmann Machines (RBMs) and autoencoders have been used - in several variants - for ... more Restricted Boltzmann Machines (RBMs) and autoencoders have been used - in several variants - for similar tasks, such as reducing dimensionality or extracting features from signals. Even though their structures are quite similar, they rely on different training theories. Lately, they have been largely used as building blocks in deep learning architectures that are called deep belief networks (instead of stacked RBMs) and stacked autoencoders.In light of this, the student has worked on this thesis with the aim to understand the extent of the similarities and the overall pros and cons of using either RBMs, autoencoders or denoising autoencoders in deep networks. Important characteristics are tested, such as the robustness to noise, the influence on training of the availability of data and the tendency to overtrain. The author has then dedicated part of the thesis to study how the three deep networks in exam form their deep internal representations and how similar these can be to each o...
International Journal of Production Research
Can automatically authored videos of industrial operators help other operators to learn procedura... more Can automatically authored videos of industrial operators help other operators to learn procedural tasks? This question is relevant to the advent of the industrial internet of things (IIoT) and Industry 4.0, where smart machines can help human operators rather than replacing them in order to benefit from the best of humans and machines. This study considers an industrial ecosystem where procedural knowledge (PK) is quickly and effectively transferred from one operator to another. Assembly tasks are procedural in nature and present a certain complexity that still does not allow machines and their sensors to capture all the details of the operations. Especially if the assembly operation is adaptive and not fixed in terms of assembly sequence plan. In order to help the operators, videos of other operators executing the complex procedural tasks can be automatically recorded and authored from machines. This study shows by means of statistical design and analysis of experiments that expert aid can reduce the assembly time of an untrained operator, whereas automatically authored video aids can transfer PK but producing an opposite effect on the assembly time. Therefore, hybrid training methods are still necessary and trade-offs have to be considered. Managerial insights from the results suggest an unneglectable impact of the choice to digitise industrial operations too early. The experimental studies presented can act as guidelines for the correct statistical testing of innovative solutions in industry.
IEEE Access, 2020
Questo articolo scientifico sostiene la tesi che un algoritmo di controllo, facente uso di intell... more Questo articolo scientifico sostiene la tesi che un algoritmo di controllo, facente uso di intelligenza artificiale, per poter essere efficiente necessiti di sfruttare le simmetrie integrate in ogni manipolatore robotico industrale così che quest’ultimo possa essere ulteriormente caratterizzato ed utilizzato. Il prodotto di questo miglioramento è uno spazio discreto a griglia cilindrica quadridimensionale (4D) che può direttamente sostituire complessi modelli robotici. A* è scelto per il suo ampio utilizzo tra simili algoritmi di ricerca, in modo da studiare i vantaggi e gli svantaggi del controllo di robot industriali tramite la griglia discreta cilindrica 4D. Lo studio mostra che questo approccio consente di controllare un robot senza alcuna conoscenza specifica dei modelli cinematici e dinamici del robot al momento della pianificazione e dell’esecuzione. In effetti, le posizioni dei giunti del robot per ciascuna cella della griglia vengono precalcolate e memorizzate come conoscenza, quindi recuperate rapidamente dall’algoritmo di ricerca di percorso quando necessario. Lo spazio discreto cilindrico 4D presenta sia i vantaggi dello spazio di configurazione che dello spazio di lavoro cartesiano tridimensionale del robot. Poiché l’ottimizzazione del percorso è il nucleo di qualsiasi algortimo di ricerca, incluso A*, la griglia cilindrica 4D fornisce uno spazio di ricerca che può incorporare ulteriori conoscenze sotto forma di proprietà delle celle, inclusa la presenza di ostacoli e l’occupazione volumetrica dell’intero corpo del robot industriale, da usare in applicazioni per l’evitamento degli ostacoli. Il compromesso principale è tra una capacità limitata della conosenza precalcolata nella griglia e la velocità di ricerca del percorso migliore. Questo approccio innovativo incoraggia l’uso di algoritmi di ricerca per applicazioni robotiche industriali, apre la via allo studio di altre simmetrie fisiche presenti in altri modelli di robot e pone le basi per l’applicazione di algoritmi dinamici per l’evitamento degli ostacoli.
Procedia Manufacturing, 2020
In the age of digital manufacturing there is a need to elicit and transfer procedural knowledge b... more In the age of digital manufacturing there is a need to elicit and transfer procedural knowledge between humans and machines. Having proper knowledge is essential in decision-making. The more the knowledge, the better decisions are made. To capture experiences and turn them into knowledge is fundamental in learning processes and knowledge development. Knowledge engineering and knowledge management have been subject for research for decades and several concepts about knowledge and knowledge transfer have been introduced, but a functional approach to exploit knowledge efficiently in manufacturing is still missing. In the era of Industry 4.0, humans and machines must be able to collaborate in such way that both can exploit the best abilities of each other in a manufacturing process. This paper introduces a procedural knowledge process (PKP) approach to capturing and defining unexpected events, while a process step is able to perform its required functions and transfer that information as machine-understandable knowledge about a failure mode. Function blocks (FBs), as per the IEC-61499 standard, have been proposed as a way to achieve distributed process planning in which the manufacturing process can adapt itself to runtime conditions, e.g. machine availability, etc. However, FBs are event-driven systems and the approach is limited to work under well-known runtime conditions, e.g. machine configurations and states, or deviations which are impossible to foresee in advance, for instance the outcome of a process failure mode effects analysis (PFMEA). The PKP introduced in this paper, aims at bridging this gap by integrating at runtime an expert operator's solution based on root cause analysis (RCA) in an FB architecture, making the FB knowledge-driven systems, for further executions of the same without redesigning it. Natural language representations of procedural knowledge blocks (PKBs) allow to transfer procedural knowledge to human operators, i.e. explain the process flow of a machine decision, while machine representations of PKBs allow to embed procedural knowledge that is elicited from expert operators upon unexpected events into the FBs process. The resulting PKP enhances the FBs for smart industrial applications, such as the process planning use case described in this paper.
Uploads
Papers by Andrea de Giorgio