Piloted flight involves the participation of both an air vehicle and a pilot. While the study of aircraft has been ongoing since the inception of aeronautical sciences, the realm of pilots and, particularly, the relationship between pilots and aircraft, the so-called Pilot-Vehicle System (PVS), is less explored. This is primarily due to the multidisciplinary nature of the problem, which touches upon disciplines far removed from aeronautical engineering and the physical sciences, such as psychology and philosophy. However, given the importance of the topic and its increasing centrality due to the growing integration of automatic systems interacting with the pilot at ever deeper cognitive levels, limiting the analysis to strict engineering treatment is not a viable approach. Human beings, in their entirety, are highly complex systems at both the cognitive and behavioural levels as well as regarding the physical body. The close relationship between the body and the behavioural-cognitive level and their mutual interaction can determine the difference between the complete success or failure of the mission The greatest complexity regarding the task of piloting, and generally any task presented to a human being, lies in extreme adaptability: control strategies rarely resemble one another between tasks, and the transition from one strategy to another can be rapid and difficult to perceive even by the operators themselves. Moreover, each operator constructs and implements strategies in a virtually unique manner, influenced by immeasurable factors such as biological predispositions, cultural situations, and acquired habits. In aviation, this latter fact is well known, as efforts have always been made to uniform the way pilots approach aircraft through standardized training delivered similarly both in fixed-wing and rotary-wing contexts; however, peculiarities remain, especially in military contexts, where differences in training arise due to the high specialization required for high-risk missions and complex scenarios. Regarding the human body, the difficulty stems from the high variability of the parameters that characterize each individual, inevitably shifting the analysis toward a statistical approach with all the difficulties it entails. Furthermore, not all parameters are easily measurable; factors such as muscle activation, tendon stiffness, or hormonal physiological reactions can be crucial in determining the response to external mechanical stimuli without being identifiable experimentally in the context of the mission situation to be analyzed. This research aims tooffer a systematic approach to analyzing pilot behaviour during a mission. Given the multifaceted nature of the problem, the primary focus is to investigatethe pilot's reaction modes to an unexpected event that causes a change in the task. The research approach proceeds through three parts: first, a meta-model for human behaviour during task performance is developed based on various psychoanalytic theories. Then, a bespoke experimental apparatus is designed to measure Control-Unrelated Inputs (CUIs) utilizing biodynamic feedthrough (BDFT) measurements. Finally, an experiment involving helicopter pilots is conducted to extract information on their susceptibility to accelerations while performing a helicopter task. The development of the pilot behaviour meta-model was preceded by a careful definition of the required terminology starting from the important assumption that human behaviour is inherently teleological, therefore directed towards a goal that can be explicit or implicit. The definitions of: · voluntariness: the state in which an act or an action is goal-driven. · Consciousness: a human's awareness of something external or internal to itself. Once the boundaries within which to operate are established, the model is developed based on an analogy with a computer where the main novelty introduced is the presence of a process scheduler responsible for context switching: tasks are managed based on priority queues. The transition between tasks is a particularly delicate operation that, despite the human predisposition to carry it out, can have unexpected consequences. The initial motivation for building the model was the interpretation of unwanted coupling phenomena between vehicle and pilot, the so-called Rotorcraft Pilot Couplings (RPC). In the available literature, RPCs are generally divided into voluntarily commanded, often referred to as Pilot Induced Oscillations (PIOs), and involuntarily commanded, often indicated as Pilot Augmented Oscillations (PAOs). This classification typically relies on the distinction between voluntary and involuntary inputs, often without defining their nature in a more precise manner, severing them depending on the frequency content of the input signal: 0-1 Hz for the voluntary,1-7.5 Hz. for the involuntary. In light of the previously given definitions, inputs are here categorized as Control-Motivated Inputs (CMI) and Control-Unrelated Inputs (CUI). The classification of RPC events is thus revised as: · Control-Motivated Couplings; · Control-Unrelated Couplings. Although both categories of phenomena are of great interest, this work focuses, at least in the experimental part, on Control-Unrelated Couplings. In previous studies, these events were studied as a pure consequence of pilot biomechanics; in this work, however, an attempt is made to demonstrate how pilot voluntariness has a non-negligible impact on the phenomenon's development. Such impact implies new possibilities in preventing and controlling these undesirable and potentially dangerous events. To conduct the required experiments, a bespoke platform was built, resembling a flight simulator, with actuation dynamics high enough to guarantee input across the entire bandwidth of interest 0-7.5 Hz. The platform carries a helicopter cockpit mock-up, whose peculiar characteristic is the ability to accurately replicate commands from multiple existing vehicles and to introduce new ones of relatively arbitrary characteristics. A series of tests, conducted with professional pilots and untrained individuals, here indicated as non-pilots, were carried out without simulating a real flight task, to obtain a statistical basis regarding the consequences of physical differences between individuals on the transmissibility of vibrations from the seat to the controls, using the tool of experimental identification of biodynamic feedthrough (BDFT). Then, a numerical closed-loop stability analysis was performed using numerical models of various rotorcraft whose data are available in the literature. Finally, using the methods developed and refined for the previous analyses, preliminary tests were conducted on two pilots of different expertise levels and training, evaluating the variation in transmissibility during the performance of purely helicopter tasks with associated unexpected events that caused an unplanned task change and a subsequent context switch. The results showed how humans respond to an unexpected event with a perceptible variation in muscle activation; this can occur consciously or unconsciously. In both cases, the mechanism falls within the pilot's voluntary behaviour as it is aimed at achieving a goal that changed following the event itself. The tests performed are not sufficient in number and variety to provide a statistically significant basis for drawing general conclusions, but they underline, once again, that the correct way to treat humans in a closed loop is that of a stochastic model whose variability is not only influenced by the natural difference across individuals but also by the situation involving the operator, including their cognitive abilities, and physiological and emotional reactions. The implications of this particular type of phenomena phenomenon can be summarized as follows: · An unexpected, startling event changes the mechanical impedance of the pilot, and therefore also of the pilot-vehicle system (PVS), and can be a trigger for a Control-Unrelated Coupling event. · The mechanism of muscle activation has a significant impact on transmissibility and, if trained to be used consciously, can become a prophylactic protocol regarding unwanted coupling events with aircraft dynamics. Hence, the need to implement an appropriate training syllabus to make pilots understand that their attitude toward the aircraft and the task they are performing within the mission has profound implications for their relationship with the machine itself and can determine the success or failure of the mission. In conclusion,the concept of voluntariness has been expanded including all the action concurring to accomplish a scope and its effects cannot be discarded in the analysis of any type of phenomena even those involving high-frequency bandwidth. Moreover, the proposed meta-model can be adapted and used in the analysis of complex pilot behaviour, possibly leading to new paradigms in the design of training procedures ornovel cockpit layouts. The work carried out in this thesis can be further developed by focusing on: · Extending closed-loop tests with a statistically significant number of pilots and with a more realistic flight simulation including proprioceptive and vestibular feedback. · An analysis using mathematical models to reconstruct the patterns used by pilots while performing specific tasks. · The numerical simulation of the context switching scheme in the prioritization of the activities to be performed, possibly using human body models to reconstruct their transmissibility in various situations.
Il volo pilotato coinvolge la partecipazione sia di un aeromobile che di un pilota. Sebbene lo studio degli aeromobili sia stato in corso fin dall'inizio delle scienze aeronautiche, il campo dei piloti e, in particolare, il rapporto tra piloti e aeromobili, il cosiddetto Sistema Pilota-Veicolo (SPV), è meno esplorato. Ciò è principalmente dovuto alla natura multidisciplinare del problema, che tocca discipline lontane dall'ingegneria aeronautica e dalle scienze fisiche, come la psicologia e la filosofia. Tuttavia, data l'importanza del tema e la sua crescente centralità a causa dell'integra\-zione sempre più profonda di sistemi automatici che interagiscono con il pilota a livelli cognitivi sempre più profondi, limitare l'analisi a un rigido trattamento ingegneristico non è un approccio praticabile. Gli esseri umani, nella loro interezza, sono sistemi altamente complessi sia a livello cognitivo e comportamentale che a livello fisico. La stretta relazione tra il corpo e il livello comportamentale-cognitivo e la loro reciproca interazione possono determinare la differenza tra il completo successo o il fallimento di una missione. La maggiore complessità riguardante il compito del pilotaggio, e in generale qualsiasi compito presentato a un essere umano, risiede nell'estrema adattabilità: le strategie di controllo raramente si somigliano tra loro tra i compiti, e la transizione da una strategia all'altra può essere rapida e molto difficile da percepire anche dagli operatori stessi. Inoltre, ogni operatore costruisce e implementa strategie in modo virtualmente unico, influenzato da fattori imponderabili come predisposizioni biologiche, situazioni culturali ed abitudini acquisite. Nell'aviazione, quest'ultimo fatto è ben noto, poiché sono stati sempre fatti sforzi per uniformare il modo in cui i piloti affrontano gli aeromobili attraverso una formazione standardizzata offerta in modo simile in tutti i domini; tuttavia, restano delle particolarità, specialmente in contesti militari, dove le differenze nella formazione sorgono a causa dell'alta specializzazione richiesta per missioni ad alto rischio e scenari complessi. Per quanto riguarda il corpo umano, la difficoltà deriva dall'alta variabilità dei parametri che caratterizzano ogni individuo, spostando inevitabilmente l'analisi verso un approccio statistico con tutte le difficoltà che comporta. Inoltre, non tutti i parametri sono facilmente misurabili; fattori come l'attivazione muscolare, la rigidità tendinea o le reazioni fisiologiche ormonali possono essere cruciali nel determinare la risposta agli stimoli meccanici esterni senza essere chiaramente identificabili sperimentalmente nel contesto della situazione di missione da analizzare. L'obiettivo di questa ricerca è offrire un approccio sistematico all'analisi del comportamento del pilota durante una missione. Date le molteplici sfaccettature del problema, è stato scelto di concentrarsi principalmente sulle modalità di reazione del pilota a un evento imprevisto che causa un cambiamento nel compito. L'approccio alla ricerca procedette in tre parti: innanzitutto, è stato sviluppato un meta-modello per il comportamento umano durante l'esecuzione di un compito basato su varie teorie psicoanalitiche. Successivamente, è stato sviluppato un apparecchio sperimentale su misura per la misurazione degli Input non correlati al Controllo (CUI) utilizzando misurazioni di biodynamica (BDFT). Infine, è stata condotta un'esperimento coinvolgente piloti di elicottero per estrarre informazioni sulla loro suscettibilità alle accelerazioni durante l'esecuzione di un compito in elicottero. Lo sviluppo del meta-modello del comportamento del pilota è stato preceduto dalla definizione attenta dei termini; partendo dall'importante assunzione che il comportamento umano sia intrinsecamente teleologico, quindi orientato verso un obiettivo che può essere esplicito o implicito. Le definizioni di ``volontarietà'' e ``coscienza'' sono state stabilite. Una volta stabiliti i confini entro cui operare, il modello è stato sviluppato basandosi su un'analogia con un computer dove la principale novità introdotta è la presenza di un pianificatore di processi responsabile dello switch di contesto: i compiti sono gestiti in base a code di priorità. La transizione tra compiti è un'operazione particolarmente delicata che, nonostante la predisposizione umana a portarla a termine, può avere conseguenze impreviste. L'impulso iniziale per la costruzione del modello è stata l'interpretazione di fenomeni di accoppiamento indesiderati tra macchina e pilota, i cosiddetti Accoppiamenti Pilota-Aeromobile (RPC). Nella letteratura disponibile, gli RPC sono generalmente divisi in comandati volontariamente, spesso indicati come Oscillazioni Indotte dal Pilota (PIO), e comandati involontariamente, noti come Oscillazioni Pilota-Aumentate (PAO). Questa classificazione si basa sulla distinzione tra input volontari e involontari, spesso senza definirne in modo più preciso la natura, separandoli in base al contenuto in frequenza del segnale in ingresso: 0-1 Hz per i volontari, 1-7,5 Hz per gli involontari. Alla luce delle definizioni precedenti, gli input sono distinti in: Input Motivati dal Controllo (CMI) e Input Non Correlati al Controllo (CUI). Segue la classificazione rivista degli eventi RPC come Accoppiamenti Motivati dal Controllo e Accoppiamenti Non Correlati al Controllo. Nonostante il grande interesse in entrambe le categorie di fenomeni, questo lavoro si concentra, almeno nella parte sperimentale, sugli Accoppiamenti Non Correlati al Controllo. Negli studi precedenti, questi eventi sono stati studiati come pura conseguenza della biomeccanica del pilota; in questo lavoro, tuttavia, si è cercato di dimostrare come la volontarietà del pilota abbia un impatto non trascurabile sullo sviluppo del fenomeno. Ciò implica nuove possibilità nella prevenzione e nel controllo di questi eventi indesiderati e potenzialmente pericolosi. Per condurre gli esperimenti richiesti, è stata costruita una piattaforma su misura, una sorta di simulatore di volo, con dinamiche di attuazione sufficientemente elevate per garantire input su tutta la banda di interesse. La piattaforma ospita un mock-up di cabina di pilotaggio dell'elicottero, la cui caratteristica peculiare è la capacità di replicare con precisione i comandi provenienti da macchine esistenti e non esistenti. È stata effettuata una serie di test, condotti con piloti professionisti e individui non addestrati, chiamati non piloti, senza simulare un compito di volo reale, per ottenere una base statistica riguardante le conseguenze delle differenze fisiche tra gli individui sulla trasmissibilità delle vibrazioni dal sedile ai comandi, utilizzando lo strumento di identificazione sperimentale della biodynamica (BDFT). Successivamente, è stata effettuata un'analisi di stabilità in anello chiuso numerico con varie macchine i cui dati sono disponibili in letteratura. Infine, armati dei metodi sviluppati e affinati per le analisi precedenti, sono stati condotti test preliminari utilizzando due piloti di diversa esperienza e formazione, valutando la variazione nella trasmissibilità durante l'esecuzione di compiti puramente elicottero con eventi imprevisti associati che hanno causato un cambio di compito non pianificato e un successivo switch di contesto. I risultati hanno mostrato come gli esseri umani rispondano a un evento imprevisto con una variazione percettibile nell'attivazione muscolare; ciò può avvenire consapevolmente o inconsapevolmente, in entrambi i casi, il meccanismo rientra nel comportamento volontario del pilota in quanto mira a raggiungere un obiettivo che è ragionevolmente cambiato a seguito dell'evento stesso. In conclusione, il concetto di volontarietà è stato ampliato includendo tutte le azioni finalizzate al raggiungimento di uno scopo e i cui effetti non possono essere scartati nell'analisi di qualsiasi tipo di fenomeno, anche quelli che coinvolgono un'ampia banda di frequenza. Inoltre, il meta-modello proposto può essere adattato e utilizzato nell'analisi del complesso comportamento del pilota, portando eventualmente a nuovi paradigmi nella progettazione di procedure di addestramento o nuove disposizioni della cabina di pilotaggio. Infine, il lavoro svolto in questa tesi può essere ulteriormente sviluppato concentrandosi su: l'estensione dei test in anello chiuso con un numero statisticamente significativo di piloti e con una simulazione di volo ancora più realistica che includa il feedback propriocettivo e vestibolare; l'analisi mediante modelli matematici per ricostruire i modelli utilizzati dai piloti durante l'esecuzione di compiti molto specifici e la simulazione numerica dello schema di switch di contesto nella prioritizzazione delle attività da svolgere, utilizzando eventualmente modelli del corpo umano per ricostruire la loro trasmissibilità in varie situazioni.
Comprehensive analysis of pilot vehicle interactions
MARCHESOLI, DAVIDE
2023/2024
Abstract
Piloted flight involves the participation of both an air vehicle and a pilot. While the study of aircraft has been ongoing since the inception of aeronautical sciences, the realm of pilots and, particularly, the relationship between pilots and aircraft, the so-called Pilot-Vehicle System (PVS), is less explored. This is primarily due to the multidisciplinary nature of the problem, which touches upon disciplines far removed from aeronautical engineering and the physical sciences, such as psychology and philosophy. However, given the importance of the topic and its increasing centrality due to the growing integration of automatic systems interacting with the pilot at ever deeper cognitive levels, limiting the analysis to strict engineering treatment is not a viable approach. Human beings, in their entirety, are highly complex systems at both the cognitive and behavioural levels as well as regarding the physical body. The close relationship between the body and the behavioural-cognitive level and their mutual interaction can determine the difference between the complete success or failure of the mission The greatest complexity regarding the task of piloting, and generally any task presented to a human being, lies in extreme adaptability: control strategies rarely resemble one another between tasks, and the transition from one strategy to another can be rapid and difficult to perceive even by the operators themselves. Moreover, each operator constructs and implements strategies in a virtually unique manner, influenced by immeasurable factors such as biological predispositions, cultural situations, and acquired habits. In aviation, this latter fact is well known, as efforts have always been made to uniform the way pilots approach aircraft through standardized training delivered similarly both in fixed-wing and rotary-wing contexts; however, peculiarities remain, especially in military contexts, where differences in training arise due to the high specialization required for high-risk missions and complex scenarios. Regarding the human body, the difficulty stems from the high variability of the parameters that characterize each individual, inevitably shifting the analysis toward a statistical approach with all the difficulties it entails. Furthermore, not all parameters are easily measurable; factors such as muscle activation, tendon stiffness, or hormonal physiological reactions can be crucial in determining the response to external mechanical stimuli without being identifiable experimentally in the context of the mission situation to be analyzed. This research aims tooffer a systematic approach to analyzing pilot behaviour during a mission. Given the multifaceted nature of the problem, the primary focus is to investigatethe pilot's reaction modes to an unexpected event that causes a change in the task. The research approach proceeds through three parts: first, a meta-model for human behaviour during task performance is developed based on various psychoanalytic theories. Then, a bespoke experimental apparatus is designed to measure Control-Unrelated Inputs (CUIs) utilizing biodynamic feedthrough (BDFT) measurements. Finally, an experiment involving helicopter pilots is conducted to extract information on their susceptibility to accelerations while performing a helicopter task. The development of the pilot behaviour meta-model was preceded by a careful definition of the required terminology starting from the important assumption that human behaviour is inherently teleological, therefore directed towards a goal that can be explicit or implicit. The definitions of: · voluntariness: the state in which an act or an action is goal-driven. · Consciousness: a human's awareness of something external or internal to itself. Once the boundaries within which to operate are established, the model is developed based on an analogy with a computer where the main novelty introduced is the presence of a process scheduler responsible for context switching: tasks are managed based on priority queues. The transition between tasks is a particularly delicate operation that, despite the human predisposition to carry it out, can have unexpected consequences. The initial motivation for building the model was the interpretation of unwanted coupling phenomena between vehicle and pilot, the so-called Rotorcraft Pilot Couplings (RPC). In the available literature, RPCs are generally divided into voluntarily commanded, often referred to as Pilot Induced Oscillations (PIOs), and involuntarily commanded, often indicated as Pilot Augmented Oscillations (PAOs). This classification typically relies on the distinction between voluntary and involuntary inputs, often without defining their nature in a more precise manner, severing them depending on the frequency content of the input signal: 0-1 Hz for the voluntary,1-7.5 Hz. for the involuntary. In light of the previously given definitions, inputs are here categorized as Control-Motivated Inputs (CMI) and Control-Unrelated Inputs (CUI). The classification of RPC events is thus revised as: · Control-Motivated Couplings; · Control-Unrelated Couplings. Although both categories of phenomena are of great interest, this work focuses, at least in the experimental part, on Control-Unrelated Couplings. In previous studies, these events were studied as a pure consequence of pilot biomechanics; in this work, however, an attempt is made to demonstrate how pilot voluntariness has a non-negligible impact on the phenomenon's development. Such impact implies new possibilities in preventing and controlling these undesirable and potentially dangerous events. To conduct the required experiments, a bespoke platform was built, resembling a flight simulator, with actuation dynamics high enough to guarantee input across the entire bandwidth of interest 0-7.5 Hz. The platform carries a helicopter cockpit mock-up, whose peculiar characteristic is the ability to accurately replicate commands from multiple existing vehicles and to introduce new ones of relatively arbitrary characteristics. A series of tests, conducted with professional pilots and untrained individuals, here indicated as non-pilots, were carried out without simulating a real flight task, to obtain a statistical basis regarding the consequences of physical differences between individuals on the transmissibility of vibrations from the seat to the controls, using the tool of experimental identification of biodynamic feedthrough (BDFT). Then, a numerical closed-loop stability analysis was performed using numerical models of various rotorcraft whose data are available in the literature. Finally, using the methods developed and refined for the previous analyses, preliminary tests were conducted on two pilots of different expertise levels and training, evaluating the variation in transmissibility during the performance of purely helicopter tasks with associated unexpected events that caused an unplanned task change and a subsequent context switch. The results showed how humans respond to an unexpected event with a perceptible variation in muscle activation; this can occur consciously or unconsciously. In both cases, the mechanism falls within the pilot's voluntary behaviour as it is aimed at achieving a goal that changed following the event itself. The tests performed are not sufficient in number and variety to provide a statistically significant basis for drawing general conclusions, but they underline, once again, that the correct way to treat humans in a closed loop is that of a stochastic model whose variability is not only influenced by the natural difference across individuals but also by the situation involving the operator, including their cognitive abilities, and physiological and emotional reactions. The implications of this particular type of phenomena phenomenon can be summarized as follows: · An unexpected, startling event changes the mechanical impedance of the pilot, and therefore also of the pilot-vehicle system (PVS), and can be a trigger for a Control-Unrelated Coupling event. · The mechanism of muscle activation has a significant impact on transmissibility and, if trained to be used consciously, can become a prophylactic protocol regarding unwanted coupling events with aircraft dynamics. Hence, the need to implement an appropriate training syllabus to make pilots understand that their attitude toward the aircraft and the task they are performing within the mission has profound implications for their relationship with the machine itself and can determine the success or failure of the mission. In conclusion,the concept of voluntariness has been expanded including all the action concurring to accomplish a scope and its effects cannot be discarded in the analysis of any type of phenomena even those involving high-frequency bandwidth. Moreover, the proposed meta-model can be adapted and used in the analysis of complex pilot behaviour, possibly leading to new paradigms in the design of training procedures ornovel cockpit layouts. The work carried out in this thesis can be further developed by focusing on: · Extending closed-loop tests with a statistically significant number of pilots and with a more realistic flight simulation including proprioceptive and vestibular feedback. · An analysis using mathematical models to reconstruct the patterns used by pilots while performing specific tasks. · The numerical simulation of the context switching scheme in the prioritization of the activities to be performed, possibly using human body models to reconstruct their transmissibility in various situations.| File | Dimensione | Formato | |
|---|---|---|---|
|
Marchesoli_PhD_Final.pdf
accessibile in internet per tutti
Descrizione: PhD thesis
Dimensione
29.61 MB
Formato
Adobe PDF
|
29.61 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/228632