Refine
Year of publication
- 2019 (18) (remove)
Document Type
Language
- English (18)
Has Fulltext
- yes (18)
Is part of the Bibliography
- no (18) (remove)
Keywords
- Maschinelles Lernen (3)
- machine learning (3)
- Künstliche Intelligenz (2)
- Personenkraftwagen (2)
- Rückenschmerz (2)
- Sport (2)
- back pain (2)
- sports (2)
- Aguascalientes (Staat) (1)
- Analyse (1)
Institute
- FB Umweltplanung/-technik (UCB) (7)
- FB Informatik + Therapiewissenschaft (5)
- ISS - Institut für Softwaresysteme in Wirtschaft, Umwelt und Verwaltung (2)
- FB Bauen + Leben (1)
- FB Wirtschaft (1)
- IfaS - Institut für angewandtes Stoffstrommanagement (1)
- LaROS - Labor für Radiotechnologie und optische Systeme (1)
This article presents experience curves and cost-benefit analyses for electric and plug-in hybrid cars sold in Germany. We find that between 2010 and 2016, the prices and price differentials relative to conventional cars declined at learning rates of 23 ± 2% and 32 ± 2% for electric cars and 6 ± 1% and 37 ± 2% for plug-in hybrids. If trends persist, price beak-even with conventional cars may be reached after another 7 ± 1 million electric cars and 5 ± 1 million plug-in hybrids are produced. The user costs of electric and plug-in hybrid cars relative to their conventional counterparts are declining annually by 14% and 26%. Also the costs for mitigating CO2 and air pollutant emissions through the deployment of electrified cars tend to decline. However, at current levels, NOX and particle emissions are still mitigated at lower costs by state-of-the-art after-treatment systems than through the electrification of powertrains. Overall, the observation of robust technological learning suggests policy makers should focus their support on non-cost market barriers for the electrification of road transport, addressing specifically the availability of recharging infrastructure.
Passenger cars in Europe have become both heavier and more powerful over the past decades. This trend has increased vehicle utility but it might have also offset technical improvements in powertrain efficiency. Here, we analyze efficiency trade-offs and CO2 emissions for three popular compact cars in Germany. We find that mass, power, and front area of model variants has increased by 66%, 147%, and 22%, respectively between 1980 and 2018. In the same period, fuel consumption decreased 14% for gasoline models but it increased 9% for diesel models. However, if vehicle mass, power, and front area had remained at 1980 levels, technical efficiency improvements would have decreased the fuel consumption of gasoline and diesel models by 23% and 24%, respectively. The related efficiency trade-offs amount to 24 g CO2/km or 13% of the current fuel consumption for gasoline models and 40 g CO2/km or 25% of the current fuel consumption for diesel models. These findings suggest that about half of the technical efficiency improvements in gasoline models and all of the technical efficiency improvements in diesel models are offset through other vehicle attributes. By accounting for the observed efficiency trade-offs, climate policy could become more effective.
We present the concrete realization of a virtual laboratory equipped with a pedagogical agent. Its functionality and media didactics takes into account the results of an usability test on a prototype system, and the students' demand on such an automated assistance as obtained from a preliminary survey. The pedagogical agent mediates between the content and the learner by activating him or her. To provide information about the learner's skills, we propose a pragmatic and simplified competence model that is based on fundamental representations in physics (experiment, figure, text and equation). Moreover, an automated feedback relates the student's self-assessment with the submitted answer to the correctness of the respective task. In consequence, the pedagogical agent enables mental reflection for a crucial review of the own learning process. Interestingly, learning pathways can be envisioned, thus, giving valuable insight into individual strengths and weaknesses.
Introduction: Injury prevention programs (IPPs) are an inherent part of training in recreational and professional sports. Providing performance-enhancing benefits in addition to injury prevention may help adjust coaches and athletes’ attitudes towards implementation of injury prevention into daily routine. Conventional thinking by players and coaches alike seems to suggest that IPPs need to be specific to one’s sport to allow for performance enhancement. The systematic literature review aims to firstly determine the IPPs nature of exercises and whether they are specific to the sport or based on general conditioning. Secondly, can they demonstrate whether general, sports-specific or even mixed IPPs improve key performance indicators with the aim to better facilitate long-term implementation of these programs?
Methods: PubMed and Web of Science were electronically searched throughout March 2018. The inclusion criteria were randomized control trials, publication dates between Jan 2006 and Feb 2018, athletes (11–45 years), injury prevention programs and included predefined performance measures that could be categorized into balance, power, strength, speed/agility and endurance. The methodological quality of included articles was assessed with the Cochrane Collaboration assessment tools.
Results: Of 6619 initial findings, 22 studies met the inclusion criteria. In addition, reference lists unearthed a further 6 studies, making a total of 28. Nine studies used sports specific IPPs, eleven general and eight mixed prevention strategies. Overall, general programs ranged from 29–57% in their effectiveness across performance outcomes. Mixed IPPs improved in 80% balance outcomes but only 20–44% in others. Sports-specific programs led to larger scale improvements in balance (66%), power (83%), strength (75%), and speed/agility (62%).
Conclusion: Sports-specific IPPs have the strongest influence on most performance indices based on the significant improvement versus control groups. Other factors such as intensity, technical execution and compliance should be accounted for in future investigations in addition to exercise modality.
Background: High numbers of consumable medical materials (eg, sterile needles and swabs) are used during the daily routine of intensive care units (ICUs) worldwide. Although medical consumables largely contribute to total ICU hospital expenditure, many hospitals do not track the individual use of materials. Current tracking solutions meeting the specific requirements of the medical environment, like barcodes or radio frequency identification, require specialized material preparation and high infrastructure investment. This impedes the accurate prediction of consumption, leads to high storage maintenance costs caused by large inventories, and hinders scientific work due to inaccurate documentation. Thus, new cost-effective and contactless methods for object detection are urgently needed.
Objective: The goal of this work was to develop and evaluate a contactless visual recognition system for tracking medical consumable materials in ICUs using a deep learning approach on a distributed client-server architecture.
Methods: We developed Consumabot, a novel client-server optical recognition system for medical consumables, based on the convolutional neural network model MobileNet implemented in Tensorflow. The software was designed to run on single-board computer platforms as a detection unit. The system was trained to recognize 20 different materials in the ICU, while 100 sample images of each consumable material were provided. We assessed the top-1 recognition rates in the context of different real-world ICU settings: materials presented to the system without visual obstruction, 50% covered materials, and scenarios of multiple items. We further performed an analysis of variance with repeated measures to quantify the effect of adverse real-world circumstances.
Results: Consumabot reached a >99% reliability of recognition after about 60 steps of training and 150 steps of validation. A desirable low cross entropy of <0.03 was reached for the training set after about 100 iteration steps and after 170 steps for the validation set. The system showed a high top-1 mean recognition accuracy in a real-world scenario of 0.85 (SD 0.11) for objects presented to the system without visual obstruction. Recognition accuracy was lower, but still acceptable, in scenarios where the objects were 50% covered (P<.001; mean recognition accuracy 0.71; SD 0.13) or multiple objects of the target group were present (P=.01; mean recognition accuracy 0.78; SD 0.11), compared to a nonobstructed view. The approach met the criteria of absence of explicit labeling (eg, barcodes, radio frequency labeling) while maintaining a high standard for quality and hygiene with minimal consumption of resources (eg, cost, time, training, and computational power).
Conclusions: Using a convolutional neural network architecture, Consumabot consistently achieved good results in the classification of consumables and thus is a feasible way to recognize and register medical consumables directly to a hospital’s electronic health record. The system shows limitations when the materials are partially covered, therefore identifying characteristics of the consumables are not presented to the system. Further development of the assessment in different medical circumstances is needed.
The increasing availability of off-the-shelf high-frequency components makes radar measurement become popular in mainstream industrial applications. We present a cooperative FM radar for strongly reflective environments, being devised for a range of up to approx. 120 m. The target is designed with an unambiguous signature method and satisfies coherence. A prototype is built with commercial semiconductor components that operates in the 24 GHz industrial, scientific and medical band. First experimental results taken in sewage pipes are presented, using the target prototype and a standard FMCW radio station. An overview on four data acquisition procedures is given.
Online Learning algorithms and Indoor Positioning Systems are complex applications in the environment of cyber-physical systems. These distributed systems are created by networking intelligent machines and autonomous robots on the Internet of Things using embedded systems that enable the exchange of information at any time. This information is processed by Machine Learning algorithms to make decisions about current developments in production or to influence logistics processes for optimization purposes. In this article, we present and categorize the further development of the prototype of a novel Indoor Positioning System, which constantly adapts its knowledge to the conditions of its environment with the help of Online Learning. Here, we apply Online Learning algorithms in the field of sound-based indoor localization with low-cost hardware and demonstrate the improvement of the system over its predecessor and its adaptability for different applications in an experimental case study.
Background: Stratified care is an up-to-date treatment approach suggested for patients with back pain in several guidelines. A comprehensively studied stratification instrument is the STarT Back Tool (SBT). It was developed to stratify patients with back pain into three subgroups, according to their risk of persistent disabling symptoms. The primary aim was to analyse the disability differences in patients with back pain 12 months after inclusion according to the subgroups determined at baseline using the German version of the SBT (STarT-G). Moreover, the potential to improve prognosis for disability by adding further predictor variables, an analysis for differences in pain intensity according to the STarT-Classification, and discriminative ability were investigated.
Methods: Data from the control group of a randomized controlled trial were analysed. Trial participants were members of a private medical insurance with a minimum age of 18 and indicated as having persistent back pain. Measurements were made for the risk of back pain chronification using the STarT-G, disability (as primary outcome) and back pain intensity with the Chronic Pain Grade Scale (CPGS), health-related quality of life with the SF-12, psychological distress with the Patient Health Questionnaire-4 (PHQ-4) and physical activity. Analysis of variance (ANOVA), multiple linear regression, and area under the curve (AUC) analysis were conducted.
Results: The mean age of the 294 participants was 53.5 (SD 8.7) years, and 38% were female. The ANOVA for disability and pain showed significant differences (p < 0.01) among the risk groups at 12 months. Post hoc Tukey tests revealed significant differences among all three risk groups for every comparison for both outcomes. AUC for STarT-G’s ability to discriminate reference standard ‘cases’ for chronic pain status at 12 months was 0.79. A prognostic model including the STarT-Classification, the variables global health, and disability at baseline explained 45% of the variance in disability at 12 months.
Conclusions: Disability differences in patients with back pain after a period of 12 months are in accordance with the subgroups determined using the STarT-G at baseline. Results should be confirmed in a study developed with the primary aim to investigate those differences.
Fuzzy system based on two-step cascade genetic optimization strategy for tobacco tar prediction
(2019)
There are many challenges in accurately measuring cigarette tar constituents. These include the need for standardized smoke generation methods related to unstable mixtures. In this research were developed algorithms using fusion of artificial intelligence methods to predict tar concentration. Outputs of development are three fuzzy structures optimized with genetic algorithms resulting in genetic algorithm (GA)-FUZZY, GA-adaptive neuro fuzzy inference system (ANFIS), GA-GA-FUZZY algorithms. Proposed algorithms are used for the tar prediction in the cigarette production process. The results of prediction are compared with gas chromatograph (high-performance liquid chromatography (HPLC)) readings.
Background: To facilitate access to evidence-based care for back pain, a German private medical insurance offered a health program proactively to their members. Feasibility and long-term efficacy of this approach were evaluated.
Methods: Using Zelen’s design, adult members of the health insurance with chronic back pain according to billing data were randomized to the intervention (IG) or the control group (CG). Participants allocated to the IG were invited to participate in the comprehensive health program comprising medical exercise therapy and life style coaching, and those allocated to the CG to a longitudinal back pain survey. Primary outcomes were back pain severity (Korff’s Chronic Pain Grade Questionnaire) as well as health-related quality of life (SF-12) assessed by identical online questionnaires at baseline and 2-year follow-up in both study arms. In addition to analyses of covariance, a subgroup analysis explored the heterogeneity of treatment effects among different risks of back pain chronification (STarT Back Tool).
Results: Out of 3462 persons selected, randomized and thereafter contacted, 552 agreed to participate. At the 24-month follow-up, data on 189 of 258 (73.3%) of the IG were available, in the CG on 255 of 294 (86.7%). Significant, small beneficial effects were seen in primary outcomes: Compared to the CG, the IG reported less disability (1.6 vs 2.0; p = 0.025; d = 0.24) and scored better at the SF-12 physical health scale (43.3 vs 41.0; p < 0.007; d = 0.26). No effect was seen in back pain intensity and in the SF-12 mental health scale. Persons with medium or high risk of back pain chronification at baseline responded better to the health program in all primary outcomes than the subgroup with low risk at baseline.
Conclusions: After 2 years, the proactive health program resulted in small positive long-term improvements. Using risk screening prior to inclusion in the health program might increase the percentage of participants deriving benefits from it.