Improving R&D processes by an ISO 9001-based Quality Management System

This paper presents practical experience in planning, implementing, and adopting a Quality Management System (QMS) for embedded systems development at VTT Electronics. The main objective for the development of the QMS was to make it practical for real-life embedded systems research and development (R&D) projects. We have applied the ISO 9001 standard and ISO 9000–3 guidelines in building the quality system. From our personnel's point of view, the most important parts of the system have been document skeletons and plan templates that were made accessible to everyone. One of the major quality improvement resulting from the use of the QMS has been the enhanced predictability of R&D projects. This makes it possible for the organization to concentrate on essential matters. From our clients' point of view, we have clearly improved the quality of our R&D services in terms of the customer satisfaction index. Based on QMS auditing activities, we have been able to evaluate and tune our R&D procedures in a systematic manner. We have decided to use Total Quality Management and Quality Award frameworks in the further development of the QMS. Customer service and human resources management are examples of important areas for further quality improvement.     

Training linear ranking SVMs in linearithmic time using red-black trees

We introduce an efficient method for training the linear ranking support vector machine. The method combines cutting plane optimization with red-black tree based approach to subgradient calculations, and has O(ms + mlog (m)) time complexity, where m is the number of training examples, and s the average number of non-zero features per example. Best previously known training algorithms achieve the same efficiency only for restricted special cases, whereas the proposed approach allows any real valued utility scores in the training data. Experiments demonstrate the superior scalability of the proposed approach, when compared to the fastest existing RankSVM implementations.     

AUTOMATED OBJECT LOCATOR IN A FOREST FORWARDER

A double approach method for versatile detection of similar objects from the real environment is introduced. By searching both the shape and textural data from the video image, it is possible to solve a generalized location problem without a loss in overall reliability in near–real time. A forest forwarder is used as a proof-of-concept application for the technique.     

Too good to be bad: Favorable product expectations boost subjective usability ratings

In an experiment conducted to study the effects of product expectations on subjective usability ratings, participants (N = 36) read a positive or a negative product review for a novel mobile device before a usability test, while the control group read nothing. In the test, half of the users performed easy tasks, and the other half hard ones, with the device. A standard usability test procedure was utilized in which objective task performance measurements as well as subjective post-task and post-experiment usability questionnaires were deployed. The study revealed a surprisingly strong effect of positive expectations on subjective post-experiment ratings: the participants who had read the positive review gave the device significantly better post-experiment ratings than did the negative-prime and no-prime groups. This boosting effect of the positive prime held even in the hard task condition where the users failed in most of the tasks. This finding highlights the importance of understanding: (1) what kinds of product expectations participants bring with them to the test, (2) how well these expectations represent those of the intended user population, and (3) how the test situation itself influences and may bias these expectations.     

Habits make smartphone use more pervasive

Examining several sources of data on smartphone use, this paper presents evidence for the popular conjecture that mobile devices are “habit-forming.” The form of habits we identified is called a checking habit: brief, repetitive inspection of dynamic content quickly accessible on the device. We describe findings on kinds and frequencies of checking behaviors in three studies. We found that checking habits occasionally spur users to do other things with the device and may increase usage overall. Data from a controlled field experiment show that checking behaviors emerge and are reinforced by informational “rewards” that are very quickly accessible. Qualitative data suggest that although repetitive habitual use is frequent, it is experienced more as an annoyance than an addiction. We conclude that supporting habit-formation is an opportunity for making smartphones more “personal” and “pervasive.”     

Neural network modelling of word production in Finnish: coding semantic and non-semantic features

The objective of our research is to computationally model word production and its disorders by means of artificial neural networks. In the current study we develop and analyze an algorithm that generates a distributed semantic coding from a given semantic tree-structure classification of words. With the algorithm it is possible to generate semantic representations that are compact and easy to modify. This renders the coding method suitable for our multilayer perceptron-based neural network model of word production. The model is shown to be able to account for a variety of performance patterns observed in four Finnish aphasia patients suffering from word-finding difficulties.     

What the small Rubik’s cube taught me about data structures, information theory, and randomisation

This article discusses observations made when the state space of the 2 × 2 × 2 Rubik’s cube was constructed with various programs based on various data structures, gives theoretical explanations for the observations, and uses them to develop more memory-efficient data structures. The cube has 3,674,160 reachable states. The fastest program runs in 20 s and uses 11.1 million bytes of memory for the state set structure. It uses a 31-bit representation of the state and also stores the rotation through which each state was first found. Its memory consumption is remarkably small, considering that 3,674,160 × 31 bits is about 14.2 million bytes. Getting below this number was made possible by sharing common parts of states. Obviously, it is not possible to reduce memory consumption without limit. We derive an information-theoretic hard average lower bound of 6.07 million bytes that applies in this setting. We introduce a general-purpose variant of the data structure and end up with 8.9 million bytes and 48 s. We also discuss the performance of BDDs and perfect state packing in this application.     

NMR investigation of the magnetic susceptibility anisotropy in3He-A

We have investigated transverse continuous wave NMR in a geometry where the modes are influenced by susceptibility anisotropy. The helium sample is separated by mylar foils into layers 12 m thick perpendicular to the external dc magnetic field. In this geometryl is constrained to be parallel to the field andd switches from parallel to perpendicular to the field over a small range of field around the dipole unlocking field. The mode frequencies below this critical field yield a value for the susceptibility anisotropy. Above the critical field the expected negative frequency shift is observed, as well as a positive frequency shift due to bulk material and an unexpected signal close to the Larmor frequency.