Customers
AutoEdge's products have served multiple customers from a wide array of applications. We help our customers use their current big data innovatively to create more reliable products and better serve their clients.
Manufacturing Company
The manufacturing industry is under constant pressure to detect defects as soon as possible in their production process. Defects in manufacturing that go undetected become more and more expensive as they advance in the manufacturing line. An error becomes the costliest when a defective product is sold to end consumers.
AutoEdge helps manufacturers quickly detect defects and provide clear actional items to fix inefficiencies.
10
Data samples required to build a comprehensive synthetic dataset
3-steps
to go from raw data to real-time monitoring
10 hrs
Warning before reaching production failure
Edge IoT
DNNs are powerful image-processing tools, but their deployment on edge devices is challenging due to limited computational resources and memory. A new solution from AutoEdge integrates DNN inference with a camera and an LCD display for image acquisition and detection exhibition, respectively.
300 KB
Total DNN model memory
91.9 ms
Inference time
1.845 mJ
Energy consumption
HVAC Company
Given its end-to-end, automated, and interpretable characteristics, AutoEdge significantly simplifies the development and deployment of AIoT-enabled remote anomaly identification while allowing domain experts to quickly understand the issues of HVAC systems. These characteristics translate to benefits, including identifying energy savings opportunities, extending asset life, minimizing repair costs, and reducing emissions.
1,000
Units analyzed simultaneously
30%+
Energy savings from preventive maintenance
25%
Fewer emissions
Oil and gas company
Given the sensor log of the oil drilling station, AutoEdge provided the attribute and physical meaning of individual variables. Specifically, with the physical attributes of the monitored values, AutoEdgre provided a typical value range and the causal relationship between variables, e.g., when pressure increases, the temperature will increase afterward.
48 hr
Warning provided before breakdown
10,000+
Outliers detect with no human intervention
24
Unique variables monitored and correlated
Video Analytics (sports)
AutoEdge leverages the power of computer vision systems by using activity tracking technologies to further the emergence of a new era in sports analytics. Our team has used player-tracking variables and game statistics to gather more data and provide relevant information about players' performance
2.5 hrs
Of video analyzed per second
1:1 score
Developed a talent-similarity score
100
Players analyzed