Beyond structural test, the rising need for system-level test

The introduction of System Level Test (SLT) about a decade ago aimed to a better sustainability for the achievement of the quality objectives required by high performances GPUs and CPUs. The paper intends giving some quantitative information to support this simplified statement. To do this we will report the results of a study conducted by the industrial application of SLT on a product targeted to a high-quality market segment, such as automotive. The discussion will describe the different SLT solutions developed for this new field of application and how it was possible to isolate SLT-only fails and relative functional root causes and cross-correlate traditional ATE test versus SLT screening capability through manufacturing operations.

To achieve high quality of silicon ICs, system-level test (SLT) can be performed after regular final test. This is important for chips manufactured in advanced technologies, as systematic failures are getting harder to detect by conventional structural tests. However, due to long test time and extra human efforts, the cost for SLT is high. A possible way to replace SLT without quality loss is to identify SLT failure suspects with stress tests. In this work, we apply 60,000 structural stress test patterns to the CPU blocks of a real SOC product, using 20 stressed voltage-frequency corners. We try to identify the correlation between the stress test data and SLT-pass/fail results of the CPU blocks. By the proposed differential feature-based methodology, 32 outliers are identified, which are assumed to be CPU-fail chips. Because of the lack of exact CPU-fail chip IDs for verification, the identified chip IDs are compared with the IDs identified from previous works, which use the same data but different machine-learning features and method for the same purpose. After comparison, 30 out of a total of 33 CPU-fail suspects matched. Although this does not immediately imply that the SLT can be replaced by the structural stress tests, it shows more evidence that test data mining can be further explored for test time reduction and/or quality improvement.

Software testing is sometimes plagued with intermittently failing tests and finding the root causes of such failing tests is often difficult. This problem has been widely studied at the unit testing level for open source software, but there has been far less investigation at the system test level, particularly the testing of industrial embedded systems. This paper describes our investigation of the root causes of intermittently failing tests in the embedded systems domain, with the goal of better understanding, explaining and categorizing the underlying faults. The subject of our investigation is a currently-running industrial embedded system, along with the system level testing that was performed. We devised and used a novel metric for classifying test cases as intermittent. From more than a half million test verdicts, we identified intermittently and consistently failing tests, and identified their root causes using multiple sources. We found that about 1-3% of all test cases were intermittently failing. From analysis of the case study results and related work, we identified nine factors associated with test case intermittence. We found that a fix for a consistently failing test typically removed a larger number of failures detected by other tests than a fix for an intermittent test. We also found that more effort was usually needed to identify fixes for intermittent tests than for consistent tests. An overlap between root causes leading to intermittent and consistent tests was identified. Many root causes of intermittence are the same in industrial embedded systems and open source software. However, when comparing unit testing to system level testing, especially for embedded systems, we observed that the test environment itself is often the cause of intermittence.

Resilience is an emergent property of complex systems. Understanding resilience is critical for sustainability science, as linked social–ecological systems and the policy process that governs them have the capacity for nonlinear dynamics. The possibility of nonlinear change in these systems means that there is an inherent degree of uncertainty in social–ecological systems and the policy process. Abrupt, nonlinear change often results in social and/or ecological surprises that create tremendous challenges for environmental management. Thus, it is necessary to improve environmental management via a suite of mechanisms that have the capacity for adaptation. This paper suggests how we can move closer to achieving this goal through an overarching focus, including reformed and new law, adaptive management and adaptive governance, scenario planning, and leading indicators.

The term resilience describes stress-response patterns across scientific disciplines. In ecology, advances have been made to clearly define resilience based on underlying mechanistic assumptions. Engineering resilience (rebound) is used to describe the ability of organisms to recover from adverse conditions (disturbances), which is termed the rate of recovery. By contrast, the ecological resilience definition considers a systemic change, that is, when ecosystems reorganize into a new regime following disturbance. Under this new regime, structural and functional aspects change considerably relative to the previous regime, without recovery. In this context, resilience is an emergent property of complex systems. In the present study, we argue that both definitions and uses are appropriate in ecotoxicology, and although the differences are subtle, the implications and uses are profoundly different. We discuss resilience concepts in ecotoxicology, where the prevailing view of resilience is engineering resilience from chemical stress. Ecological resilience may also be useful for describing systemic ecological changes because of chemical stress. We present quantitative methods that allow ecotoxicologists and risk managers to assess whether an ecosystem faces an impending regime shift or whether it has already undergone such a shift. We contend that engineering and ecological resilience help to distinguish ecotoxicological responses to chemical stressors mechanistically and thus have implications for theory, policy, and application. Environ Toxicol Chem 2017;36:2574-2580. © 2017 SETAC.

Thresholds are an emergent property of complex systems and Coupled Natural Human Systems (CNH) because they indicate “tipping points” where a complicated array of social, environmental, and/or economic processes combine to substantially change a system’s state. Because of the elegance of the concept, thresholds have emerged as one of the primary tools by which socio-political systems simplify, define, and especially regulate complex environmental impacts and resource scarcity considerations. This paper derives a general framework for the use of thresholds to calculate scarcity footprints, and presents a volumetric Threshold-based Water Footprint (TWF), comparing it with the Blue Water Footprint (BWF) and the Relevant for Environmental Deficiency (RED) midpoint impact indicator. Specific findings include (a) one requires all users’ BWF to calculate an individual user’s TWF, whereas one can calculate an individual user’s BWF without other users’ data; (b) local maxima appear in the Free from Environmental Deficiency (FED) efficiency of the RED metric due to its nonlinear form; and (c) it is possible to estimate the “effective” threshold that is approximately implied by the RED water use impact metric.

Sociomaterial Enactment Drive of Business/IT Alignment: From Small Data to Big Impact

Automotive systems must reach a high reliability in their electronic components. This kind of devices must undergo several tests and stress steps discovering all possible defects that could manifest during lifetime. Burn-In (BI) is a manufacturing test phase used for screening the early life latent faults that can naturally affect a population of devices. System Level Test (SLT) is increasingly adopted as one of the final steps in the testing process of complex Systems on Chip (SoCs) mimicking the operational conditions. This paper aims at describing the motivations for and the effectiveness stemming from combining SLT with BI. The key idea leverages on the development of a new step inside the test process, which reproduces the system using SLT and places the system in the worst cases by means of the BI. Moreover, the paper analyses the required tester architecture to merge SLT and BI. Finally, an industrial case by STMicroelectronics is used to demonstrate the possible cost reduction.

This paper shows a way to define a partially-exhaustive gate transition fault model for use in catching defects that escape when using more traditional fault models. We define the gate-level transitions ATPG must create for this fault model and how this may catch un-modeled defects. Future work will analyze results of applying tests generated using this fault model against a commercial chip design.

Device failure mechanisms of today's deep sub-micron processes are not well-modeled by single stuck-at faults, and hence several advanced fault models have been proposed in the past. Gate-exhaustive fault models were proposed to exercise a gate completely and then observe the resultant response at an observable output. This paper extends the gate-exhaustive fault model to target bigger regions (a collection of gates) with the hypothesis that exercising a region with an exhaustive pattern set can yield coverage on a larger proportion of unmodeled defects. To test out this hypothesis, we use the logic proximity bridge (LPB) fault model as a surrogate for unmodeled defects and grade the region and gate exhaustive patterns against the LPB fault model to gauge their efficacy. We show that region exhaustive patterns are better at detecting untargeted LPB faults compared to patterns obtained using gate exhaustive or traditional stuck-at fault models.

Nowadays, the electricity generation industry works with complex systems. A complex system can be defined as composed of several subsystems, elements, components, or different equipments, interacting spatially and temporarily in a nonlinear way. A system's complexity is related to the amount of different elements, intensity of the relations, and the functions that each element carries out to reach the expected operational performance. Oil production platforms and nuclear plants are examples of complex and dynamic systems. This paper presents a proposal based on the application of non homogenous Markovian Processes for availability estimation and for the decision making related to the design modification of a complex system. In this proposal the operational state are modeled according to the system failure scenario. The transition rates between those scenarios are failure rate (lambda(t)) and repair rate (mu(t)). The probabilities analysis of each operational state, calculated from the Markovian analysis process, are used as input to the decision making process. That process is based on the calculation of the expected mean cost of each decision alternative. The method is applied for the decision making on installing a water-supply redundant system of heat recovery steam generator used in a combined cycle thermal power plant operating in Brazil.

System-based testing of a voltage differential protection scheme for shunt capacitor bank

Hydraulic fracturing experiences were gathered over the years, from a tight sandstone reservoir of Late Carboniferous to Early Permian origin to find the root causes for the persisting challenges of high formation breakdown pressures and low fluid injectivity. It was soon realized that a single root cause could not account for these challenges. Instead, a multitude of reasons acting at the same time during the injecting sequence of that particular fracturing stage could easily be responsible, and no single re-occurring pattern would explicitly repeat in the consequent stages under similar conditions. It was found that the known complex nature of the depositional environment of the targeted formation played an important role, but the overall hydraulic fracturing mechanics played an equally critical part in the success of the fracturing attempt. Having realized this complexity in deciphering hydraulic fracturing performances it became imminent that not only the fracturing execution had to be refined, but all the preceding activities from drilling operations, to well placement, to wellbore preparation, to the choice of completion equipment and its deployment. These operations critically influenced the near wellbore region. The drilling fluid properties, wellbore hydraulics, well trajectory, the wellbore clean-out procedures, and hole quality were important contributors in the geo-mechanics domain coupled with the impact of the geological setting. Thus, a workflow was developed to analyze and rationalize the root causes for the injectivity challenges with the goal of optimizing every well construction step that would enhance hydraulic fracturing and ultimately deliver optimum well productivity.

There is nothing wrong with the concept of a root cause.

Failure Mode and Effects Analysis (FMEA) is a key safety assessment analysis that determine failure modes at system, hardware and software level. Overlooking Failure Modes can often cause system or functionality failure which directly impacts a systems safety performance, reliability and quality. FMEA is a bottom-up approach which has four key phases - identification of fault, assessment of impact, determination of potential causes and their resolutions, and finally testing and documentation of analysis. FMEA addresses the effect of failures at the system, software and hardware level. The outcome of the analysis helps us identify gaps in safety requirements specification and provides input for component testing, integration testing and system level testing. This paper describes the application of Failure Mode and Effects Analysis (FMEA) to software modules.