Double ambidexterity: How a Telco incumbent used business‐model and technology innovations to successfully respond to three major disruptions

We leverage the business model innovation and ambidexterity literature to investigate a contradictory case, the Swedish‐Finnish Telecom operator TeliaSonera. Despite being challenged by three major disruptions, the company not only still exists but also enjoys remarkably good financial performance. Building on extant archival data and interviews, we carefully identify and map 26 organizational responses during 1992–2016. We find that the firm has overcome three critical phases by experimenting and pioneering with portfolios of business models and/or technological innovations. We describe this behaviour as double ambidexterity. We use an in‐depth case study to conceptualize double ambidexterity and discuss its impact on the business's survival and enduring success.     

Language Data,Ethics, and the Law

This article explores some legal, ethical, and organizational difficulties facing Swedish data curators when curating voice-based language data. These problems are traced through three cases where researchers wanted their data curated by SND. Two of the materials were subsequently published in different ways by SND, while the third was rejected. The last section discusses some ways to handle these problems. None of them are miracle cures, but at least they may prepare language data curators for what lies ahead, or clear their thoughts on what is happening around them.     

Creating Ideas for Innovation: Effects of Organizational Distance on Knowledge Creation Processes

Innovation is to a large extent considered a social and communicative process, and input from other individuals potentially improves the generation of novel and valuable ideas also in the early stages of idea creation and development. Both colleagues inside organizations and external parties have frequently been proposed as important sources of information and knowledge within this part of the innovation process. Other contributions addressing social networks and innovation bring into focus the potentially negative effects that certain network structures may have on innovation, pointing to inconsistencies in received theory. In order to address these inconsistencies, an empirical study of ideation in a Swedish multinational firm was performed, taking into account two different knowledge creation processes – combination and in‐depth analysis – and their inter‐relationships with organizational distance between contributing individuals. Data was collected using a survey and was analysed using regression models. It was found that different levels of organizational distance correlate with different knowledge creation processes. In‐depth analysis occurred more often with employees' close colleagues, whereas the combination of existing ideas and information was more frequent in interaction with employees' close colleagues and with external parties. Both these interaction patterns were also found to be positive for the generation of patents, whereas no such relationship could be seen when individuals interacted with colleagues in other departments in the same firm. The findings have implications for theory on cognitive distance, and also suggest that management needs to facilitate different types of collaboration and networking when aiming to facilitate and support ideation, taking into consideration the type of innovation aimed for, as well as its supporting knowledge creation processes.     

Bayesian neural networks for prediction of equilibrium and time-dependent scour depth around bridge piers

The physical process of scour around bridge piers is complicated. Despite various models presented to predict the equilibrium scour depth and its time variation from the characteristics of the current and sediment, scope exists to improve the existing models or to provide alternatives to them. In this paper, a neural network technique within a Bayesian framework, is presented for the prediction of equilibrium scour depth around a bridge pier and the time variation of scour depth. The equilibrium scour depth was modeled as a function of five variables; flow depth and mean velocity, critical flow velocity, median grain diameter and pier diameter. The time variation of scour depth was also modeled in terms of equilibrium scour depth, equilibrium scour time, scour time, mean flow velocity and critical flow velocity. The Bayesian network predicted equilibrium and time-dependent scour depth much better when it was trained with the original (dimensional) scour data, rather than using a non-dimensional form of the data. The selection of water, sediment and time variables used in the models was based on conventional scour depth data analysis. The new models estimate equilibrium and time-dependent scour depth more accurately than the existing expressions. A committee model, developed by averaging the predictions of a number of individual neural network models, increased the reliability and accuracy of the predictions. A sensitivity analysis showed that pier diameter has a greater influence on equilibrium scour depth than the other independent parameters.     

Facit and the BESK Boys: Sweden's computer industry (1956-1962)

In 1956, the Swedish office equipment company Facit embarked on a plan to produce large-scale computers for the Swedish and Western European markets. The plan involved recruiting several top engineers known as the BESK Boys, who developed a computer for Facit that briefly was the world's fastest. Circumstances, however, ultimately worked against Facit.     

基于深度图像与骨骼数据的行为识别

为了充分利用深度图像与骨骼数据进行人体行为识别,提出了一种基于深度图形与骨骼数据的多特征行为识别方法。该算法的多特征包括深度运动图（DMM）特征与四方形骨骼特征（Quad）。深度图像方面,将深度图像投影到一个笛卡尔坐标系的三个平面获得深度运动图特征。骨骼数据方面,提出四方形骨骼特征,它是骨骼坐标的一种标定方式,得到的结果只与骨骼姿态有关。同时提出一种多模型概率投票的分类策略,减小了噪声数据对分类结果的影响。所提方法在MSR-Action3D和DHA数据库进行实验,实验结果表明,所提算法有着较高的识别率与良好的鲁棒性。     

基于曲率局部二值模式的深度图像手势特征提取

针对复杂环境下的深度图像手势特征提取信息冗余量大、编码不稳定等问题,提出了一种改进的基于曲率局部二值模式（LBP）的深度图像手势特征提取算法。该算法首先通过坐标转换将分割出的手势深度数据转换为点云数据;其次利用移动最小二乘法对手势点云数据进行曲面拟合;然后计算出能够更加准确描述物体三维表面几何信息特征的高斯曲率;最后利用改进的LBP均匀模式对高斯曲率数据进行编码形成特征向量。在美国手语（ASL）手势数据库上该算法的平均识别率达到了92.1%,与3D局部二值模式（3DLBP）和梯度LBP相比分别提高了18.5个百分点和13.7个百分点。实验结果表明,该算法可以区分外部轮廓相似但内部结构不同的手势,有效提高了在描述手势深度图像内部细节方面的准确性。     

Neural network and neuro-fuzzy assessments for scour depth around bridge piers

The mechanism of flow around a pier structure is so complicated that it is difficult to establish a general empirical model to provide accurate estimation for scour. Interestingly, each of the proposed empirical formula yields good results for a particular data set. Hence, in this study, alternative approaches, artificial neural networks (ANNs) and adaptive neuro-fuzzy inference system (ANFIS), are proposed to estimate the equilibrium and time-dependent scour depth with numerous reliable data base. Two ANN models, multi-layer perception using back-propagation algorithm (MLP/BP) and radial basis using orthogonal least-squares algorithm (RBF/OLS), were used. The equilibrium scour depth was modeled as a function of five variables; flow depth, mean velocity, critical flow velocity, mean grain diameter and pier diameter. The time variation of scour depth was also modeled in terms of equilibrium scour depth, equilibrium scour time, scour time, mean flow velocity and critical flow velocity. The training and testing data are selected from the experimental data of several valuable references. Numerical tests indicate that MLP/BP model provide a better prediction of scour depth than RBF/OLS and ANFIS models as well as the previous empirical approaches. Finally, sensitivity analysis shows that pier diameter has a greater influence on equilibrium scour depth than the other independent parameters.     

基于AMSR2和MODIS数据融合的雪深降尺度算法研究—以北疆地区为例

高空间分辨率雪深数据对于区域气候、水文研究具有重要的意义。利用10 km空间分辨率的AMSR2 L1B亮度温度数据,结合500 m空间分辨率的MODIS逐日无云积雪面积比例数据,发展了一种多源数据融合的空间动态降尺度雪深反演算法（SDD）。基于该算法获取了北疆地区500 m空间分辨率的雪深数据（SDDsd）,并利用研究区30个气象台站和野外实测的雪深数据对该算法反演雪深的精度进行了评估。结果表明：基于SDD方法获取的雪深数据与实测雪深数据之间的决定系数R²为0.74,均方根误差RMSE为3.47 cm;雪深反演的精度与下垫面类型密切相关,草地精度最高,城镇和建设用地次之,耕地相对较差;雪深反演的精度也会受到地形的影响,精度随坡度的增加而降低。相对于微波遥感雪深数据直接重采样结果,新的算法有效提高了浅雪区雪深反演精度,同时能更精细地描述积雪的空间分布,为理解区域气候变化、水文循环提供了可靠的数据支撑。此外,随着长时间序列全球尺度逐日无云FSC数据的生产,结合现有的长时间序列全球尺度AMSR2数据,该算法有望制备全球的降尺度雪深产品。     

利用高判别低维深度特征的SAR船只分类

为充分利用深度特征的判别信息,提高船只分类准确率,提出利用低维度高判别的深度特征进行SAR船只分类的方法。采用ImageNet数据库预训练的VGG16卷积神经网络作为特征提取器,提取船只样本的深度特征;对深度特征进行t-SNE可视化,计算每类船只的类间分离度,选择对于每类船只样本来说类间分离度最大的深度特征;对选择的深度特征进行降维,采用基于KNN的级联二分类方法进行船只分类。利用高分辨率SAR船只数据集验证该方法,实验结果表明,相比传统的船只分类方法,其分类性能有明显提高。     

Microcomputer based system for efficient use of electricity in industry

This paper describes the principles for industrial load management and an estimation of the load management potential in Swedish industry. Microcomputer based equipment for metering and monitoring consumption of electric energy is presented as well as collected data from some industries. Finally an example of an industrial load management application is presented.     

Crop monitoring in Sweden

Abstract

This paper deals with the most important Swedish agriculture crops in relation to (i) Landsat-based spectral differences, (ii) mono- and multi-temporal Landsat classification measured as producer's, user's and Kappa accuracy and (iii) Landsat classification determined as area estimation accuracy. The potential of satellite data for crop monitoring in Sweden was found to be rather low for most crops except rape seed and sugar beet. Concerning these crops results indicated a possible operational applicability.     

Fusion of range and color images for denoising and resolution enhancement with a non-local filter

We present an integrated method for post-processing of range data which removes outliers, smoothes the depth values and enhances the lateral resolution in order to achieve visually pleasing 3D models from low-cost depth sensors with additional (registered) color images. The algorithm is based on the non-local principle and adapts the original NL-Means formulation to the characteristics of typical depth data. Explicitly handling outliers in the sensor data, our denoising approach achieves unbiased reconstructions from error-prone input data. Taking intra-patch similarity into account, we reconstruct strong discontinuities without disturbing artifacts and preserve fine detail structures, obtaining piece-wise smooth depth maps. Furthermore, we exploit the dependencies of the depth data with additionally available color information and increase the lateral resolution of the depth maps. We finally discuss how to parallelize the algorithm in order to achieve fast processing times that are adequate for post-processing of data from fast depth sensors such as time-of-flight cameras.     

基于多源数据的阿勒泰地区雪深反演研究

利用阿勒泰地区 2010~2012年冬季(11月~次年2月)3类积雪数据:风云三号微波成像仪(FY\|3/MWRI)反演的雪深数据、美国人机交互式多仪器冰雪制图系统(IMS)积雪面积数据、阿勒泰及周边地区实测雪深数据,进行积雪深度的反演研究。通过结合3类积雪数据的各自优势,建立修正模型,最终得到较准确的研究区雪深数据。同时通过编程实现了相应模型的操作平台,为今后研究区积雪业务化监测做好准备。结果表明:模型提高了FY\|3/MWRI数据反演阿勒泰地区积雪深度的准确性,改善了FY\|3/MWRI数据在阿勒泰地区雪深反演偏低的缺点,使微波与实测平均雪深误差由修正前的21.7~12.1 cm缩小为修正后的3.7~1.5 cm。     

Prediction and optimization of depth of penetration for stainless steel gas tungsten arc welded plates using artificial neural networks and simulated annealing algorithm

The quality of a weld joint is highly influenced by depth of penetration. Hence, accurate prediction and maximization of depth of penetration is highly essential to ensure a good-quality joint. This paper highlights the development of neural network model for predicting depth of penetration and optimizing the process parameters for maximizing depth of penetration using simulated annealing algorithm. The process parameters chosen for the study are welding current, welding speed, gas flow rate and welding gun angle. The chosen output parameter was depth of penetration. The experiments were conducted based on design of experiments using fractional factorial with 125 runs. Using the experimental data, feed-forward backpropagation neural network model was developed and trained using Levenberg–Marquardt algorithm. It was found that ANN model based on network 4-15-1 predicted depth of penetration more accurately. A mathematical model was also developed correlating the process parameters with depth of penetration for doing optimization. A source code was developed in MATLAB to do the optimization. The optimized process parameters gave a value of 3.778 mm for depth of penetration.     

Combining optical and acoustic data to enhance the detection of Caribbean forereef habitats

Maps of coral reef habitats are fundamental tools for reef management, and high map accuracy is desirable to support appropriate decisions, such as the stratification of marine reserves by habitat class. While satellite sensors have been used to map different reef communities, the accuracy of these maps tends to be low (overall accuracy < 50%) and optical airborne methods with high spectral resolution have, to date, been the most effective (if expensive) means of achieving higher accuracy. A potential means of compensating for the low spectral and radiometric resolution of optical satellite data, which is a major cause of its poor performance, is to combine satellite data with acoustic remote sensing. This study quantified the benefit of the synergy between optical satellite data (IKONOS) and acoustic (RoxAnn) sensors. The addition of acoustic data provided three new data axes for discriminating habitats: seabed roughness (E1), reef depth (z) and the depth correction of satellite spectral data to uniform depth. Seabed hardness (E2) was not an informative channel in our study. The use of z to conduct the water-column correction of the optical bands to uniform depth is a potential improvement over applying the depth-invariant index approach to optical data in the absence of ancillary information on depth. Habitat maps of the forereef of Glovers Atoll (Belize, Central America) were created using k-means unsupervised classification on eleven different treatment images constructed from various combinations of optical and acoustic data layers. The maximum benefit of data synergy was achieved by depth correcting the optical bands. The accuracy of maps based on the depth-invariant optical index was not enhanced when E1, E2 or z were added as separate layers but was enhanced when these three acoustic measures were added in concert. Data synergy can improve the accuracy of habitat maps and the availability of both data sets allows practitioners to take advantage of each techniques' additional strengths such as providing synoptic continuous imagery for education and general management planning (in the case of optical imagery) and maps of reef rugosity (in the case of acoustic data).     

Limitations of the Swedish network coordination of oil spill preparedness

The management of extreme crisis situations in Sweden has shown delays in decisive actions at higher levels during emergencies, such as oil spills. This has been blamed on unclear responsibilities that undermine the decision‐making process. Confusing, conflicting, or delayed orders impede response effectiveness. This article uses network analysis and survey responses to examine the Swedish oil spill crisis management network and show the Swedish Coast Guard, Swedish Agency for Marine and Water Management, Swedish Civil Contingencies Agency and Oil Spill Advisory Service as central organizations. However, the roles of these organizations need to be clarified and communicated better to other organizations. In order to improve coordination of Swedish oil spill preparedness, it is suggested to formalize and strengthen the roles of these central organizations. Weak connections between municipalities in different counties were also observed. This weakness could be overcome by more frequent exercises across counties to improve familiarity.     

深度学习单目深度估计研究进展

单目深度估计是从单幅图像中获取场景深度信息的重要技术,在智能汽车和机器人定位等领域应用广泛,具有重要的研究价值。随着深度学习技术的发展,涌现出许多基于深度学习的单目深度估计研究,单目深度估计性能也取得了很大进展。本文按照单目深度估计模型采用的训练数据的类型,从3个方面综述了近年来基于深度学习的单目深度估计方法:基于单图像训练的模型、基于多图像训练的模型和基于辅助信息优化训练的单目深度估计模型。同时,本文在综述了单目深度估计研究常用数据集和性能指标基础上,对经典的单目深度估计模型进行了性能比较分析。以单幅图像作为训练数据的模型具有网络结构简单的特点,但泛化性能较差。采用多图像训练的深度估计网络有更强的泛化性,但网络的参数量大、网络收敛速度慢、训练耗时长。引入辅助信息的深度估计网络的深度估计精度得到了进一步提升,但辅助信息的引入会造成网络结构复杂、收敛速度慢等问题。单目深度估计研究还存在许多的难题和挑战。利用多图像输入中包含的潜在信息和特定领域的约束信息,来提高单目深度估计的性能,逐渐成为了单目深度估计研究的趋势。     

Organization and Decision‐Making in Enforced Networks: The River Groups in Northern Sweden

This article examines the Swedish experience of network governance in managing flooding and high water flows. The aim was to study the regional responsibility for coordinating risk awareness and risk analysis in terms of information, prevention and actions. The focus was on differences between the Swedish river groups from the coordinators perspective, including their organization and approaches to decision‐making. The conclusions reached here are based on interviews with the coordinators of county administrative boards. We argue that the absence of central guidelines in the organization of the river groups and the fact that they are enforced by the government rather than spontaneously formed have had implications for the networks’ effectiveness and for exchanges of experience among the networks.