Managing workload allocation to ensure fairness and equity amongst staff can be a challenge in any organisation and managing the workload allocation of autonomy seeking academic staff in a business school can be even more so. In this study, the researcher aimed to review a recently designed and implemented academic workload allocation model in a South African business school in order to establish whether the model and implementation system has been successful in contributing to actual and perceived fairness and equity in workload distribution amongst their academic staff. The researcher did this by using a sequential exploratory mixed methods approach, first reviewing documentary evidence, which informed the design of an online survey with the academic staff, followed by semi-structured interviews with a sample group. The study reveals that the model, and the way it was implemented and managed, failed to achieve its intended aims of increased equitable and fair workloads amongst academic staff. These implementation failures have resulted in negative consequences for the organisational culture. Staff satisfaction and engagement with the model, its implementation and management does not present positively in the findings of this study. In the South African context where there are very few studies related to academic workload allocation models, the results of this study may be valuable for higher education institutions considering the introduction or review of workload models amongst their academic staff. The study highlights the importance of an inclusive and careful design approach, change management considerations during the implementation phase, and the transparent management of the workload allocation process and results.

No description available.

The proliferation of the Internet of Things (IoT) has increasingly demanded intimacy between cloud services and end-users. This has incentivised extending cloud resources to the edge in what is deemed fog computing. The latter is manifesting as an ecosystem of connected clouds, geo-dispersed and of diverse capacities. In such conditions, workload allocation to fog services becomes a non-trivial challenge due to the complexity of trade-offs. Users' demand at the edge is highly diverse, which does not lend itself to straightforward resource planning. Conversely, running services at the edge may leverage proximity, but it comes at higher operational cost let alone rapidly increasing the risk of straining sparse resources. Consequently, there is a need for intelligent yet scalable allocation solutions that counter the adversity of demand at the edge, while efficiently distributing load between the edge and farther clouds. Machine learning is increasingly adopted in resource planning. However, besides privacy concerns, central learning is highly demanding, both computationally and in data supply. Instead, this paper proposes a federated deep reinforcement learning system, based on deep Q-learning network (DQN), for workload distribution in a fog ecosystem. The proposed solution adapts a DQN to optimize local workload allocations, made by single gateways. Federated learning is incorporated to allow multiple gateways in a network to collaboratively build knowledge of users' demand. This is leveraged to establish consensus on the fraction of workload allocated to different fog nodes, using lower data supply and computation resources. The system performance is evaluated using realistic demand set from Google Cluster Workload Traces 2019. Evaluation results show over 50% reduction in failed allocations when distributing users over larger number of gateways, given fixed number of fog nodes. The results further illustrate the trade-offs between performance and cost under different conditions.

L'IDentification par Radio-Fréquences (RFID) permet une identification rapide et efficace des objets. Dans les systèmes de personnalisation de masse, elle promet un apport considérable grâce à sa capacité à gérer des flux d'information complexes caractérisant ce type de systèmes.Dans cette thèse, nous étudions les impacts de la RFID sur les systèmes de configuration à la demande (CTO). Nous nous basons sur un cas réel pour apporter une mesure fiable et directement exploitable. La littérature à ce sujet offre souvent des mesures sommaires, théoriques ou qualitatives, d'où l'originalité de la thèse.L'étude est réalisée par une approche de simulation à évènements discrets et évalue l'apport des technologies RFID à deux niveaux. Le premier concerne des changements directs du système (e.g. accélération des maintes vérifications caractérisant la CTO, libération de ressources...). Ces changements influencent la performance du système en termes de temps de séjour, de taux de retard des commandes, etc. Le deuxième niveau concerne des changements profonds tirant profit d'une visibilité accrue des produits et de la facilité d'une collecte de données rigoureuse. Ces changements se focalisent sur l'allocation dynamique de la charge de travail. La remise en question des processus à l'occasion de l'introduction d'une technologie RFID constitue un point original en raison du manque de publications soulignantsuffisamment cet avantage.Nos expérimentations ont montré que les apports des technologies RFID dans un système CTO sont indéniables. De plus, repenser le fonctionnement du système afin d'exploiter plus profondément le potentiel de la technologie accroit les bénéfices. / Radio Frequency IDentification allows quick and secure identification of objects. In mass customisation systems, RFID technologies can be peculiarly efficient, because they are able to support the complex flows of information which characterize these systems.In this study, we focus on RFID technologies effects on configuration to order (CTO) systems.We base the research on an existing case in order to obtain reliable information directly usable by decision makers. The rarity of studies offering quantitative, detailed and real case based measures makes the originality of this thesis.RFID technology implementation's effect is analysed by a discrete event simulation approach and is presented in two levels:The first level relates direct changes brought about by RFID (e.g. faster execution of the many checks due to the wide range of products, reduced workload for resources…). These changes have an impact on system's performance in terms of lead time, late orders' rate, etc.The second level is axed on deeper changes occurring due to the increased product visibility and the ease of collecting large amounts of data with an RFID technology.These changes mainly focus on the dynamic allocation of workload. Reconsidering of processes and proposing changes deeper than the simple direct technology impact is a breakthrough, in this study, because of the lack of publications highlighting this benefit adequately.In conclusion, RFID contribution in CTO systems and, extensively, in assembly to order systems may be undeniable. Moreover, beyond the direct technology impact, rethinking how the system works by exploiting the deeper potential of technology can increase profits.