Search for: All records

As edge computing and sensing devices continue to proliferate, distributed machine learning (ML) inference pipelines are becoming popular for enabling low-latency, real-time decision-making at scale. However, the geographically dispersed and often resource-constrained nature of edge devices makes them susceptible to various failures, such as hardware malfunctions, network disruptions, and device overloading. These edge failures can significantly affect the performance and availability of inference pipelines and the sensing-to-decision-making loops they enable. In addition, the complexity of task dependencies amplifies the difficulty of maintaining performant and reliable ML operations. To address these challenges and minimize the impact of edge failures on inference pipelines, this paper presents several fault-tolerant approaches, including sensing redundancy, structural resilience, failover replication, and pipeline reconfiguration. For each approach, we explain the key techniques and highlight their effectiveness and tradeoffs. Finally, we discuss the challenges associated with these approaches and outline future directions. 

Abstract Climate warming in combination with nutrient enrichment can greatly promote phytoplankton proliferation and blooms in eutrophic waters. Lake Taihu, China, is a large, shallow and eutrophic system. Since 2007, this lake has experienced extensive nutrient input reductions aimed at controlling cyanobacterial blooms. However, intense cyanobacterial blooms have persisted through 2017 with a record‐setting bloom occurring in May 2017. Causal analysis suggested that this bloom was sygenerically driven by high external loading from flooding in 2016 in the Taihu catchment and a notable warmer winter during 2016/2017. High precipitation during 2016 was associated with a strong 2015/2016 El Niño in combination with the joint effects of Atlantic Multi‐decadal Oscillation (AMO) and Pacific Decadal Oscillation (PDO), while persistent warmth during 2016/2017 was strongly related to warm phases of AMO and PDO. The 2017 blooms elevated water column pH and led to dissolved oxygen depletion near the sediment, both of which mobilized phosphorus from the sediment to overlying water, further promoting cyanobacterial blooms. Our finding indicates that regional climate anomalies exacerbated eutrophication via a positive feedback mechanism, by intensifying internal nutrient cycling and aggravating cyanobacterial blooms. In light of global expansion of eutrophication and blooms, especially in large, shallow and eutrophic lakes, these regional effects of climate anomalies are nested within larger scale global warming predicted to continue in the foreseeable future.