Cyber-security is a rising issue for automotive electronic systems, and it is critical to system safety and dependability. Current in-vehicles architectures, such as those based on the Controller Area Network (CAN), do not provide direct support for secure communications. When retrofitting these architectures with security mechanisms, a major challenge is to ensure that system safety will not be hindered, given the limited computation and communication resources. We apply Message Authentication Codes (MACs) to protect against masquerade and replay attacks on CAN networks, and propose an optimal Mixed Integer Linear Programming (MILP) formulation for solving the mapping problem from a functional model to the CAN-based platform while meeting both the security and the safety requirements. We also develop an efficient heuristic for the mapping problem under security and safety constraints. To the best of our knowledge, this is the first work to address security and safety in an integrated formulation in the design automation of automotive electronic systems. Experimental results of an industrial case study show the effectiveness of our approach. © 2013 IEEE.

Can we create engaging training programs that improve working memory (WM) skills? While there are numerous procedures that attempt to do so, there is a great deal of controversy regarding their efficacy. Nonetheless, recent meta-analytic evidence shows consistent improvements across studies on lab-based tasks generalizing beyond the specific training effects (Au et al., 2014; Karbach and Verhaeghen, 2014), however, there is little research into how WM training aids participants in their daily life. Here we propose that incorporating design principles from the fields of Perceptual Learning (PL) and Computer Science might augment the efficacy of WM training, and ultimately lead to greater learning and transfer. In particular, the field of PL has identified numerous mechanisms (including attention, reinforcement, multisensory facilitation and multi-stimulus training) that promote brain plasticity. Also, computer science has made great progress in the scientific approach to game design that can be used to create engaging environments for learning. We suggest that approaches integrating knowledge across these fields may lead to a more effective WM interventions and better reflect real world conditions.

Routing-based synthesis for digital microfluidic biochips yields faster assay execution times compared to module-based synthesis. We show that routing-based synthesis can lead to deadlocks and livelocks in specific cases, and that dynamically detecting them and adjusting the probabilities associated with different droplet movements can alleviate the situation. We also introduce methods to improve the efficiency of wash droplet routing during routing-based synthesis, and to support nonreconfigurable modules, such as integrated heaters and detectors. We obtain increases in success rates when dealing with resource-constrained chips and reductions in average assay execution time.

This paper describes a publicly available, open source software framework designed to support research efforts on algorithms and control for digital microfluidic biochips (DMFBs), an emerging laboratory-on-a-chip (LoC) technology. The framework consists of two parts: a compiler, which converts an assay, specified using the BioCoder language, into a sequence of electrode activations that execute out the assay on the DMFB; and a printed circuit board (PCB) layout tool, which includes algorithms to reduce the number of control pins and PCB layers required to drive the chip from an external source. The framework also includes a suite of visualization tools for debugging, and a collection of front-end algorithms that generate mixing/dilution trees for sample preparation.

In the current literature, there are a number of cognitive training studies that use N-back tasks as their training vehicle; however, the interventions are often bland, and many studies suffer from considerable attrition rates. An increasingly common approach to increase participant engagement has been the implementation of motivational features in training tasks; yet, the effects of such "gamification" on learning have been inconsistent. To shed more light on those issues, here, we report the results of a training study conducted at two Universities in Southern California. A total of 115 participants completed 4 weeks (20 sessions) of N-back training in the laboratory. We varied the amount of "gamification" and the motivational features that might make the training more engaging and, potentially, more effective. Thus, 47 participants trained on a basic color/identity N-back version with no motivational features, whereas 68 participants trained on a gamified version that translated the basic mechanics of the N-back task into an engaging 3D space-themed "collection" game (Deveau et al. Frontiers in Systems Neuroscience, 8, 243, 2015). Both versions used similar adaptive algorithms to increase the difficulty level as participants became more proficient. Participants' self-reports indicated that the group who trained on the gamified version enjoyed the intervention more than the group who trained on the non-gamified version. Furthermore, the participants who trained on the gamified version exerted more effort and also improved more during training. However, despite the differential training effects, there were no significant group differences in any of the outcome measures at post-test, suggesting that the inclusion of motivational features neither substantially benefited nor hurt broader learning. Overall, our findings provide guidelines for task implementation to optimally target participants' interest and engagement to promote learning, which may lead to broader adoption and adherence of cognitive training.