Over the Air · 16 Feb 2023 · 30m
Josh Becker, Senior Global Director and Head of IoT Commercial at SAS, discusses the evolution of analytics from batch processing to real-time AI-driven insights. He shares his background in reliability engineering at Sub-Zero Wolf, where he built one of the earliest IoT devices to monitor refrigerators remotely. Becker explains how SAS leverages event stream processing and computer vision to solve enterprise problems, from defect detection in manufacturing to predicting utility asset failures through partnerships like the one with Exactor.
Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.
Quality & test
AI detects defects in manufacturing panels down to a pixel size, allowing for automatic quality inspection without human intervention.
“We've deployed computer vision in multiple manufacturing environments where, uh, if you think of a panel, like a panel in your TV or a panel, uh, solar panel, we can detect defects automatically using computer vision down to a pixel size…”
Speaker 1 · Hear it at 10:07
Engineering & simulationSASover 90% accuracy
AI models trained on satellite images autonomously monitor and detect deforestation with over 90% accuracy, eliminating the need for humans in the loop.
“So this is like crowd trained, deep learning AI models that can now monitor and detect deforestation autonomously without humans in the loop, with over 90% accuracy.”
Speaker 1 · Hear it at 10:07
Engineering & simulationSAS
AI models replicate the abilities of indigenous trackers to identify species, age, sex, and direction from footprint images, enabling non-invasive wildlife monitoring.
“And what we've done is work with them to train models that can replicate the work and the abilities of indigenous trackers. So indigenous trackers can look at the footprints and understand the species, whether it's a lion or a cheetah, the age, the sex, the direction…”
Speaker 1 · Hear it at 10:07
large enterprises
Other shows on the same ground, closest first.
Alessandro Fossato, Co-Founder and CEO of Interlogica, discusses the evolution of IoT software from simple connectivity to complex data-driven business models. He explains how Interlogica shifted from agile IT consulting to building custom IoT solutions that bridge the gap between physical sensors and executive decision-making. Fossato highlights the importance of edge computing and machine learning for processing large data volumes, using a real-world example of AI-based quality assurance in manufacturing.
Ira Cohen, co-founder and Chief Data Scientist of Anodot, explains how his company uses autonomous machine learning to detect anomalies in time-series data for Industrial IoT and business monitoring. He details the shift from static thresholds to dynamic pattern recognition, citing use cases in automotive manufacturing and large-scale printing to illustrate predictive maintenance and quality control. Cohen also discusses the challenges of implementing AI analytics at scale, including the need for organizational process changes and the importance of user-friendly interfaces for non-technical executives.
Steve Liang, Founder and CTO of SensorUp, explains how his company evolved from academic research on IoT interoperability to building a platform that connects siloed industrial systems for better data visibility and prediction. He details SensorUp's pivot into the 'connected worker' space, driven by the need to empower frontline employees in dangerous environments like methane leak detection, and discusses the critical importance of standardizing data exchange for climate tech applications. Liang also shares his 'AI and IoT Canvas' framework, advising companies to prioritize prediction tasks that offer competitive advantage over simple monitoring.
Elvira Wallis, Senior Vice President and Global Head of IoT at SAP, discusses how SAP integrates real-world industrial data into business processes using its IoT portfolio, including SAP Leonardo IoT and SAP Edge Services. She highlights use cases with partners like Pregis, Shifa, and Excitel, demonstrating how IoT enables predictive maintenance, inventory optimization, and new business models like 'power as a service.' Wallis emphasizes that successful IoT deployments require clear ROI, tight time-boxing in proof-of-concept stages, and strong ecosystem partnerships involving hardware vendors and system integrators.
Rohit Tripathi, SVP and Head of Products at SAP Digital Interconnect, discusses the evolution of IoT from simple device connectivity to intelligent enterprise solutions that integrate people and processes. He highlights SAP's products IoT Connect 365 and People Connect 365, explaining how blending sensor data with HR and operational systems improves safety and maintenance efficiency. Tripathi reflects on 2019's key trends, including the rollout of 5G, the maturity of cloud providers like AWS and Azure, and the shift toward solution-based ROI, while predicting greater standardization and adoption in 2020.
The summary, chapters and topics are generated from the episode’s audio by this directory, not written by Over the Air. The episode itself belongs to its publisher.