The power utility of South Africa, Eskom, expected a supply shortfall of approximately 400MW between February and August 2006 in the Western Cape. The peak of the crisis was in mid-winter (June to August). This shortfall was firstly caused when Eskom experienced a breakdown on the one of the two nuclear supply units. Secondly the remaining of the Koeberg units was due for refuelling which necessitated the shut-down of the reactor. No electricity was therefore generated by both units. It was clear that if electricity demand was not effectively curbed, extensive power outages would be experienced; which was the case. Various demand side management (DSM) programmes were rolled-out to address lighting, switching from electricity to gas for cooking, compensating customers that could generate own electricity, energy efficiency and load curtailment in the education, commercial, and industrial sectors, as well as an extensive energy efficiency campaign. It is shown in this study that the most constrained periods were expected during the evening peak and was a consequence of electricity consumption in the residential sector. The residential evening peak is very prominent and primarily caused by water heating, cooking, space heating, lighting, and appliances. None of the mentioned programmes focused on the residential evening peak. Traditional residential DSM technologies were almost impossible to implement in the short timeframe because there are more than 625,000 residences in the Western Cape. A solution was looked for that could be implemented in a relatively short period to address the residential evening peak. This study focuses on the development, implementation, and performance evaluation of Power Alert – An innovative residential load management system. The need for such a system was identified and the expected impact was determined through a feasibility study. Power Alert was designed to be a link between Eskom and the public through the national television broadcaster. It was operational during the whole Western Cape winter. A methodology to determine the impact of Power Alert was also developed to demonstrate the actual load reductions. The methodology was applied and Power Alert demonstrated that it was a valuable residential load management tool that could be designed and implemented in a much shorter time than conventional residential DSM measures. / Thesis (Ph.D. (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2010.

The power utility of South Africa, Eskom, expected a supply shortfall of approximately 400MW between February and August 2006 in the Western Cape. The peak of the crisis was in mid-winter (June to August). This shortfall was firstly caused when Eskom experienced a breakdown on the one of the two nuclear supply units. Secondly the remaining of the Koeberg units was due for refuelling which necessitated the shut-down of the reactor. No electricity was therefore generated by both units. It was clear that if electricity demand was not effectively curbed, extensive power outages would be experienced; which was the case. Various demand side management (DSM) programmes were rolled-out to address lighting, switching from electricity to gas for cooking, compensating customers that could generate own electricity, energy efficiency and load curtailment in the education, commercial, and industrial sectors, as well as an extensive energy efficiency campaign. It is shown in this study that the most constrained periods were expected during the evening peak and was a consequence of electricity consumption in the residential sector. The residential evening peak is very prominent and primarily caused by water heating, cooking, space heating, lighting, and appliances. None of the mentioned programmes focused on the residential evening peak. Traditional residential DSM technologies were almost impossible to implement in the short timeframe because there are more than 625,000 residences in the Western Cape. A solution was looked for that could be implemented in a relatively short period to address the residential evening peak. This study focuses on the development, implementation, and performance evaluation of Power Alert – An innovative residential load management system. The need for such a system was identified and the expected impact was determined through a feasibility study. Power Alert was designed to be a link between Eskom and the public through the national television broadcaster. It was operational during the whole Western Cape winter. A methodology to determine the impact of Power Alert was also developed to demonstrate the actual load reductions. The methodology was applied and Power Alert demonstrated that it was a valuable residential load management tool that could be designed and implemented in a much shorter time than conventional residential DSM measures. / Thesis (Ph.D. (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2010.

With the high-level developments in the information technologies, there is a big movement in the e-health area both to give better healthcare services and to reduce cost. Monitoring the patients from their home location is one of the important branches of e-health with the aging societies. In this work, an agentbased alert system is introduced which analyzes the alert message requests and sends the alert messages within the appropriate structure and communication channel to the related healthcare user. A guideline execution system produces the alert requests based on sensor data and informs the alert agents. A multi-agent platform is developed which proactively sends the alert messages on behalf of the system with acknowledgement and message routing abilities. The multi-agent platform has patient alert agents for each patient in the system. A rule engine is cooperated with agents for the analysis of the alert message requests to bind these requests with the healthcare user specifications. E-mail messaging, SMS and MSN Instant messaging are used to reach the healthcare users. Additionally, web-based user interfaces are developed for healthcare users to manage the alert system for intelligent healthcare monitoring by specifying alert message receiving rules, personal and contact information, and some more additional options. The work presented in this thesis is realized as a part of the SAPHIRE project funded by the European Commission.

Gamma-ray bursts (GRBs) are short, intense flashes of 0.1-1 MeV electromagnetic radiation that are routinely observed by Earth orbiting satellites. The sources of GRBs are known to be extragalacitic and located at cosmological distances. Due to the extremely high isotropic equivalent energies of GRBs, which are on the order of Eiso~1054 erg, the gamma-ray emission is believed to be collimated, making them observable only when they are directed towards Earth. The favored progenitor models of GRBs are also believed to emit gravitational waves that would be observable by the current generation of ground-based interferometric gravitational wave detectors. The LIGO (Laser Interferometer Gravitational-Wave Observatory) and Virgo instruments operated near design sensitivity and collected more than a year of triple coincident data during the S5/VSR1 science run, which spanned the two year interval between November 2005 and October 2007. During this time, GRB detections were being made by the NASA/Goddard Swift Burst Alert Telescope at a rate of approximately 0.3 per day, producing a collection of triggers that has since been used in a coincident GRB-GW burst search with data from the LIGO-Virgo interferometer network. This dissertation describes the search for gravitational waves using the times and locations of 123 below-threshold potential GRB triggers from Swift over the same time period. Although most of the below-threshold triggers are likely false alarms, there is reason to believe that some are the result of actual faintly-observed GRB events. Recent GRB observations indicate that the local rate of low-luminosity GRBs is much higher than previously believed. This result, combined with the possibility of discovering a rare nearby GRB event accompanied by gravitational waves, is what motivates this search. The analysis results indicate no evidence for gravitational waves associated with any of the below-threshold triggers. A median distance lower limit of ~16 Mpc was derived for a typical neutron star-black hole coalescence progenitor assumption.

Les besoins croissants continuels de la médecine et de la santé accentuent la nécessité de systèmes d'alertes médicales bien adaptés. Ces systèmes d'alerte peuvent être utilisés par une variété de patients et d'acteurs médicaux. Ils devraient permettre de contrôler un large éventail de variables médicales. Nous proposons Tempas, un système d'alertes paramétrées par les utilisateurs. Aux alertes détectées on associées deux indices de qualité. L'indice d'applicabilité qui exprime dans quelle mesure un patient est concerné par l'alerte, et l'indice de confiance qui exprime la fiabilité de l'alerte concernant la fraicheur des données utilisées lors de sa détection. Les indices de qualité associés à une alerte détectée se calculent grâce aux informations liées à une situation d'alerte paramétrée par l'utilisateur. Une situation d'alerte est définie à partir d'une composition de conditions d'activation. Une condition d'activation est construite à partir d'une valeur linguistique exprimant l'état (par exemple température élevée) ou la tendance (par exemple tension systolique à la hausse) d'une entité observable (la température, la tension systolique, etc.). Lorsque la situation d'alerte est évaluée, le système utilise les connaissances préparées préalablement par les utilisateurs concernant les valeurs linguistiques. C'est-à-dire, quelle valeur linguistique représente le mieux une valeur quantitative sous un contexte spécifique. Les alertes détectées pouvant être nombreuses, nous définissons un politique de notification pour ne notifier que les alertes pertinentes dans le but de ne pas fatiguer les utilisateurs. D'abord les alertes sont filtrées à partir des indices de qualité. Parmi les alertes restantes, le système filtre par expressivité : choisir les tendances plus durables et les valeurs linguistiques les plus expressives. Ensuite, dans le cas d'alertes consécutives, le système ne garde que les alertes qui accomplissent les préférences des utilisateurs, comme celles dont l'indice d'applicabilité augmente. L'objectif final est de fidéliser l'utilisateur au système d'alertes en lui fournissant un service de qualité. Il s'approprie le système en même temps qu'il définit les situations d'alerte. Ainsi, il est capable de l'adapter lui-même au contexte afin d'obtenir des alertes de meilleure qualité. L'adaptation est guidée par les indices de qualité utilisés pour réduire les faux-positifs et faux-négatifs ainsi que pour contrôler la sur-notification d'alertes. Nous proposons de nous appuyer sur les systèmes existants en apportant des fonctionnalités de dynamisme et d'évolution, ainsi que des facilités de paramétrage et d'adaptation en temps réel au contexte d'utilisation afin d'exploiter au mieux les observations. / The continuous increasing needs in medicine and healthcare, accentuate the need of well-adapted medical alert systems. Such alert systems may be used by a variety of patients and medical actors. These systems should allow monitoring a wide range of medical variables. We work on Tempas an alert system setup by users. Detected alerts have two quality indices. The applicability index which indicates how well a patient is affected by the alert, and the confidence index, which expresses the reliability of the alert concerning the freshness of the data used in its detection. Quality indices associated with a detected alert are calculated using information related to an alert situation configured by the user. An alert situation is defined from multiple activation conditions. An activation condition is constructed from a linguistic value expressing the state (e.g. high temperature) or trend (e.g. systolic blood pressure rising) of an observable entity (temperature, systolic blood pressure, etc.). When the alert condition is evaluated, the system uses knowledge previously prepared by users regarding linguistic values. That is, what linguistic value best represents a quantitative value in a specific context. Since many alerts can be detected, we define a notification policy to notify only the relevant alerts in order to keep the users' interest. First alerts are filtered from the quality indices. Of the remaining alerts, the system filters by expressiveness: to keep more sustainable trends and the most expressive linguistic values. Then, in the case of consecutive alerts, the system keeps only those that fulfill the user preferences, such as those whose applicability index increases. The ultimate goal is to keep the user loyalty to the alert system by providing quality service. The user appropriates the system while he defines the alert situation. Thus, he is able to adapt the alert situation by himself to the context to obtain better alerts. The adaptation is guided by quality indices used to reduce false positives and false negatives as well as to control the over-alerting. We propose to leverage existing systems by providing dynamism and evolution features, as well as facilities for setup and real-time adaptation to the context of use in order to fully exploit the observations.

The purpose of this Master’s Thesis is to produce a specification for the aircraft conflict alert system STCA, and implement a prototype as a module in the air traffic surveillance system NOVA9000. The specification is constructed based on functional requirements from EUROCONTROL and describes a system using a nominal trajectory method, where the future paths of aircraft are estimated. The trajectory is created using a probabilistic approach, where future positions are described with probability fields. The prototype is implemented using the specification with some simplifications. The prototype is evaluated using recorded traffic from a heavy air traffic region surrounding an airport with parallel runways. 15 alerts were induced in 1,5 hour of morning traffic; this is far too much to be acceptable. Improvements are proposed and explanations to the high rate of alerts are made.

A variabilidade espaçotemporal das inundações determinam os processos ecológicos e influenciam as atividades humanas no Pantanal. O entendimento da dinâmica das inundações é a chave para melhores práticas de gerenciamento e reação no caso de eventos extremos. Nesse estudo, métodos de quantificação de água e extração de sazonalidade foram aplicados numa série temporal de dez anos (2000 a 2009) das imagens MODIS13Q1 para a modelagem contínua das inundações. Foram determinadas 25 sub-regiões, totalizando 150.502 km2, baseado no regime das inundações e na geomorfologia. A área inundada variou de 5.539 km2 até 52.894 km2, com média de 18.329 km2, refletindo um período seco comparado às décadas anteriores. A maior parte do Pantanal (86.841 km2) inundou pelo menos uma vez no período, mas uma área relativamente grande permaneceu não inundada ou inundou por um período curto de tempo. Áreas não inundadas ou inundadas por curtos períodos de tempo ficaram espalhadas por todo o Pantanal, enquanto que áreas que inundam por longos períodos de tempo ou permanentemente inundadas se concentraram apenas em alguns locais. Quatro estágios principais de inundação foram determinados e os principais fluxos e armazenagem de água foram mapeados embora tenha ocorrido uma alta variabilidade interanual. Apesar das limitações de escala e problemas de sobreposição dos mapas, foi obtida uma boa concordância entre o mapa de frequência de inundação e os mapas de vegetação e solos. Modelos estatísticos logísticos não lineares foram determinados para as relações entre os principais compartimentos hidrológicos e foi possível estimar cenários passados de inundação baseado na relação entre a área inundada total do Pantanal e o nível do rio Paraguai. Relações de similaridade e de fluxos de água foram determinados entre os sub-compartimentos, caracterizando o comportamento espacial e temporal da água. Baseado nos dados e informações gerados foi proposto um sistema de alerta de inundações e secas para o Pantanal (SISMONPAN). / The flood variability in space and time determine the ecological process and influences human activities in the Pantanal wetland. The understanding of the flood dynamics is the key for better management practices and reaction in case of extreme events. In this study, methods for water quantification and seasonality extraction were applied to a ten years time series (2000 to 2009) of MODIS13Q1 images for continuous modeling of floods. Based in flood regime and geomorphology an area of 150.502 Km2 distributed in 25 sub-regions was determined for the Pantanal. The flooded area varied from 5.539 km2 to 52.894 km2, averaging 18.329km2, reflecting a dry period compared with the previous decades. Most of the Pantanal (86.841 km2) flooded at least once in the period, but a relative big area keeps no flooded or flooded in a short period. No flooded or short time flooded areas was spread over all the Pantanal, while long time to permanent flooded areas were concentrated in just some spots. Four main stages of floods were determined and the main water fluxes and water storage areas were mapped, although a high inter annual variability has been occurred. Despite the limitations of scale and map matching, a good agreement between flood frequency map with vegetation and soils maps was obtained. Nonlinear Logistic statistical models were determined for the relationships between the main hydrological compartments. It was possible to estimate past flood scenarios based in the relationship between total flooded area and Paraguay River stage. Similarity and water fluxes relationships were determined between the sub-compartments, characterizing the spatial and temporal water behavior. Based on the data and information generated, an alert system for dry and flood events (SISMONPAN) was proposed for the Pantanal.

A pesquisa teve por objetivo determinar o desenvolvimento, o número de gerações e a constante térmica de Diaphorina citri Kuwayama, 1908 (Hemiptera: Liviidae) e Tamarixia radiata (Waterston, 1922) (Hymenoptera: Eulophidae), sob 20 combinações de temperaturas alternantes que simulam as condições de Limeira, Tatuí, Araraquara, Santa Cruz do Rio Pardo e Votuporanga, áreas citrícolas do estado de São Paulo, nas quatro estações do ano. Paralelamente, discutem-se as estimativas dos períodos de liberação do parasitoide associados ao alerta fitossanitário desenvolvido pelo FUNDECITRUS. Foi estudado o desenvolvimento de D. citri e T. radiata, sendo registrado o tempo necessário para completarem o ciclo biológico. A partir destes dados, foi calculada a constante térmica e o número de gerações da praga e do parasitoide para cada uma das condições propostas. Com base na duração do desenvolvimento de D. citri, foi definido o momento ideal para liberação de T. radiata em campo. Os resultados da determinação do número de gerações de D. citri e de T. radiata, baseando-se em temperaturas alternantes, foram comparáveis ao modelo desenvolvido em temperaturas constantes, sendo que as variações existentes entre os dois modelos se deveram à inclusão do período de pré-oviposição que foi variável em função da temperatura. A previsão do número de gerações se ajustou mais a T. radiata do que a D. citri, sendo propostos fatores de correção para casos em que a constante térmica observada superou 10- 15% do valor determinado em laboratório. Observou-se que uma geração de D. citri variou de 22,9 a 74,2 dias, no verão e inverno para as simulações de Votuporanga e Tatuí, respectivamente; por outro lado, nas mesmas condições, uma geração de T. radiata pode variar de 9,8 a 21,6 dias. A relação entre o número de gerações de D. citri: T. radiata no decorrer do ano é superior a 1:2, sendo de 1: 2,58; 1: 2,47; 1: 2,28; 1: 2,28; 1: 2,34, respectivamente, para as simulações de Limeira, Tatuí, Araraquara, Santa Cruz do Rio Pardo e Votuporanga. A regressão linear mostrou que o número de gerações de D. citri é melhor explicado pela média das temperaturas diurna e noturna, enquanto que para T. radiata, apenas pela temperatura diurna. Os valores extremos de dias para liberação de T. radiata após o alerta fitossanitário, correspondem às regiões de Avaré e Bebedouro. Os resultados indicam que há uma variação no desenvolvimento da praga e do seu parasitoide entre as regiões e que há necessidade de validação em campo; no entanto, se confirmados, permitirão o uso mais eficiente do parasitoide T. radiata, se tornando um componente auxiliar ao sistema de alerta e ao Manejo Integrado dessa importante doença, o HLB. / This study determined the development, number of generations and thermal constant of the HLB vector Diaphorina citri Kuwayama, 1908 (Hemiptera: Liviidae) and the parasitoid Tamarixia radiata (Waterston, 1922) (Hymenoptera: Eulophidae) under 20 combinations of alternating temperatures that simulated the conditions of five citrus-growing areas of São Paulo state (Limeira, Tatuí, Araraquara, Santa Cruz do Rio Pardo and Votuporanga) in the four seasons. The estimated optimum release time of the parasitoid is discussed in relation to a phytosanitation alert developed by FUNDECITRUS. The development time and time needed to complete the life cycles of D. citri and T. radiata were assessed. The thermal constant and number of generations of the pest and parasitoid for each temperature condition were calculated based on these data. Based on the development time of D. citri, the optimum time to release T. radiata in the field was established. The estimated number of generations of D. citri and T. radiata in alternating temperatures was comparable to the model developed at constant temperatures; the differences between the two models were due to the inclusion of the pre-oviposition period, which varied as a function of temperature. The estimated number of generations was larger for T. radiata than D. citri, and correction factors were proposed for those cases in which the observed thermal constant exceeded 10-15% of the value determined in the laboratory. Generation times of D. citri ranged from 22.9 days in summer to 74.2 days in winter for the simulations for Votuporanga and Tatuí. Under the same conditions, one generation of T. radiata may range from 9.8 to 21.6 days. The ratio between the annual number of generations of D. citri: T. radiata was higher than 1: 2, ranging from 1: 2.58; 1: 2.47; 1: 2.28; 1: 2.28; 1: 2.34 for the simulations for Limeira, Tatuí, Araraquara, Santa Cruz do Rio Pardo and Votuporanga, respectively. Linear regression analysis showed that the number of generations of D. citri is best explained by the mean diurnal and nocturnal temperatures, whereas for T. radiata, only by the mean diurnal temperature. The extreme values for optimum release day of T. radiata after a phytosanitation alert were found for Avare and Bebedouro. The results indicate that the development of the pest and its parasitoid differ among areas, and there is need for validation in the field. However, if confirmed, these data will allow more efficient use of the parasitoid T. radiata as a complementary component of the alert system and of Integrated Management of HLB.

INTRODUÇÃO: O tema segurança do paciente ganhou importância a partir da publicação dos Estudos Harvard I e II, e dimensão pública a partir do livro To Error is Human pelo Institute of Medicine em 1999, onde são apresentados números alarmantes sobre erros durante o processo do cuidado em saúde. Esses estudos demonstram que os erros com medicamentos são a causa mais frequente de incidentes em pacientes internados. O Institute for Safe Medication Practices (ISMP), uma organização que se dedica à prevenção de erros de medicação e ao uso seguro dos medicamentos publicou em 1989 a primeira lista de medicamentos reconhecidos como perigosos. Em 1995, o ISMP avaliou os erros notificados com medicamentos quanto à gravidade e os danos causados no Medication Error Reporting and Prevention (MERP). Após este estudo, o termo \"High Alert Medication\" foi adotado para designar um grupo de medicamentos mais relacionados a danos graves ou fatais quando ocorre alguma falha no seu processo de utilização. No Brasil estes medicamentos são conhecidos como Medicamentos de Alto Risco, Medicamentos de Alta Vigilância (MAV) ou Medicamentos Potencialmente Perigosos (MPP). OBJETIVOS: Avaliar a ocorrência de incidentes com medicamentos de acordo com a classificação MPP e suas variáveis no Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo (HCFMRP - USP), nos anos de 2013 e 2015. MÉTODOS: Estudo transversal descritivo retrospectivo com abordagem quantitativa dos incidentes notificados ao Núcleo de Segurança do Paciente relacionados aos MPP. RESULTADOS: Durante o ano de 2013, foram notificados 28% de incidentes envolvendo MPP na unidade Campus e 37,5% na Unidade de Emergência (UE). Os medicamentos quimioterápicos foram os mais notificados e com as maiores Taxas de Incidência (TI) na unidade Campus; o cloreto de potássio foi o mais notificado e com a maior TI na UE. A etapa de prescrição foi a mais notificada nas duas unidades. Utilizando a classificação ATC, os subgrupos terapêuticos que mais atingiram o paciente foram Análogos da Purina, Análogos do Ácido Fólico e Nutrição Parenteral (Campus); Insulina e Análogos Injetáveis, Outras Preparações Cardíacas foram os mais notificados na UE. Em 2015, os percentuais de notificações com MPP foram 8,8% (Campus) e 31,7% (UE). Nesse ano, os medicamentos mais notificados foram o cloridrato de tramadol e a enoxaparina nas duas unidades. Na unidade Campus a etapa de dispensação foi a mais notificada, e etapa de administração na UE. Os subgrupos com maiores TI foram Agentes Alquilantes, Anti Histamínico para Uso Sistêmico e Agente com Ação no Músculo Liso Arteriolar (Campus); na UE, Antiarrítmico, Classe III e Analgésico Opióide. CONCLUSÃO: A classificação MPP pode padronizar a atuação do farmacêutico clínico, além de prover indicadores clínico-gerenciais que auxiliem no desenho de processos proativos de prevenção de erros de medicação. / INTRODUCTION: The patient safety issue gained importance from the publication of Harvard Studies I and II, and public dimension from the book To Error is Human by the Institute of Medicine in 1999, which presents the alarming dates about errors during the process of healthcare. These studies demonstrate that the errors with medications are the most common cause of incidents in hospitalized patients. The Institute for Safe Medication Practices (ISMP), an organization dedicated to the prevention of medication errors and the safe use of medication published in 1989 the first list of recognized dangerous drugs. In 1995, the ISMP evaluated the reports of drugs errors according the severity and damage to the Medication Error Reporting and Prevention (MERP). After this study, the term \"High Alert Medication\" was adopted to designate a group of drugs more related to serious injury or death occurs when a fault in its usage. In Brazil, these drugs are known as high-risk drugs, high alert medications or potentially dangerous drugs. OBJECTIVES: Analyze the occurrence of reported incidents involving the high alert medication and its variables at the Hospital of Ribeirão Preto Medical School, University of São Paulo (HCFMRP - USP) in the years 2013 and 2015. METHODS: Retrospective descriptive cross-sectional study with a quantitative approach of the reported incidents to the Patient Safety Center related to high alert medication. RESULTS: During the year 2013 it was reported 28% of incidents involving high alert medication in the Campus Unit (CU) and 37.5% at the Emergency Unit (EU). Chemotherapeutic drugs were the most reported and with the highest Incidence Rates (IR) (CU) and 19.1% potassium chloride was the most commonly reported and the largest IR in the EU. Prescription stage was the most reported in both units. Using the ATC classification, therapeutic subgroups most reached the patient were Purine Analogues, Folic Acid Analogues and Parenteral Nutrition (CU), Insulin and Analogs for Injections and Other Cardiac Preparations (EU). In 2015, the percentage of notifications with high alert medication was 8.8% (CU) and 31.7% (EU). The most reported drugs were tramadol hydrochloride and enoxaparin in both units. On CU dispensing stage was the most notified and administration stage in the EU. Subgroups with higher IR were Alkylating Agents, Antihistamine for Systemic Use and Arteriolar Smooth Muscle, Agent Action On (CU) and Antiarrhythmic, Class III and Analgesic Opioid (EU). CONCLUSION: The MPP classification can standardize the performance of the clinical pharmacist, and provide clinical and management indicators to assist in the proactive process design to prevent medication errors.

Developing and testing complex information technology (IT) systems is a difficult task. This is even more difficult if parts of the system, no matter if hard- or software, are not available when needed due to delays or other reasons. The architecture and design cannot be evaluated and existing parts cannot be reliably tested. Thus the whole concept of the system cannot be validated before all parts are in place. To solve this problem in an industrial project, where the development of the server-side should be finished and tested (clear for production) while the hardware sensors where designed but not implemented, we developed an emulator (software) for the hardware sensors meeting the exact specification as well as parts of the server solution. This allowed proceeding with the server-side development, testing, and system validation without the hardware sensors in place. Following the exact specification should allow replacing the emulator with the real sensors without complications, once they are available. In the end, being able to develop hard- and software in parallel the project can be in production much earlier than performing the development in sequence.