Brad Sweeney - Sr. Product Manager, HPE
The retail industry is changing constantly as new products flood the market and companies tackle emerging trends, rising consumer demands, shrinking margins, and huge amounts of data. Retailers need solutions that provide valuable information about shopping behaviors and increased visibility across the supply chain. Savvy companies are tapping into artificial intelligence (AI) to meet these needs.
AI identifies key insights from consumer and operational data that guide retailers to improve how they work and enhance in-store experiences. The combination of AI with applications like video analytics is critical for retailers to gain more control and a deeper understanding of their environments.
Video analytics powered by AI can give shoppers more personalized experiences, provide information to improve stocking levels, streamline merchandising floor plans with more effective traffic patterns, improve inventory management, and identify shoppers for special promotions or services. It can also reduce theft and loss—which is now a $100 billion problem. In the supply chain, video analytics can reduce operational costs and create greater efficiencies to ensure that products reach consumers as quickly as possible.
Harness real-world insights instantly
Hewlett Packard Enterprise has been collaborating with NVIDIA to empower retailers to work smarter and more effectively with a breed of AI built specifically for them. HPE offers pioneering systems, software, and pretrained models, as well as decades of experience, to unlock the full value of retail data. We know AI is the key to finding new insights and opportunities that boost profits and help maintain a competitive edge. That’s why we’ve also partnered with IronYun to turn real-time video data into better business outcomes.
IronYun, an NVIDIA Metropolis partner, adds intelligence to retail operations using AI video analytics. Retailers can make any IP camera an AI powerhouse to improve security and track goods at each stage of the supply chain. Layered on top of an accelerated compute platform, IronYun solutions deliver high accuracy and performance for a variety of use cases (i.e., temperature detection, safety monitoring, and store analytics).
The IronYun Vaidio AI Vision Platform, powered by NVIDIA GPUs, delivers 30+ advanced video analytics functions to add a layer of superhuman insight to existing camera and video infrastructure. The platform extends beyond basic security applications and supports numerous advanced AI and video analytics capabilities. The most common uses of Vaidio in retail include:
- Real-time monitoring: Monitoring dozens to hundreds to thousands of camera streams in real time with high alertness and accuracy
- Pattern recognition: Analyzing large volumes of video footage and identifying patterns that may indicate theft
- Data mining: Mining the video stream across all cameras to report on store behaviors such as dwell time, heat maps, bounce rates, and areas of higher loss for retail business intelligence and merchandising optimization
Validated solution for simple rollout
HPE collaboration with NVIDIA and IronYun offers great benefits to channel partners and end customers alike. The bundled solution has been specifically tested for vision AI deployments, which reduces the complexities included with standing up a proof of concept, identifying the most optimized system, and testing retail-specific software providers. Now, retailers can start with a tested and recommended architecture to streamline the evaluation and implementation process.
IronYun conducted detailed testing to determine how the solution would scale. The results showed that a single HPE ProLiant DL320 can support up to 157 camera streams at a 0% drop rate.* This testing also enabled IronYun to accurately size the number of servers/GPUs required to manage a given set of analytic channels. With 30+ AI engines (video analytics tools) included in Vaidio, organizations can scale their most valuable use cases confidently. Vaidio also features dynamic scheduling, which can automatically shift between different tools. For example, it would allow a retail center to run dwell–time analytics and heat mapping during the day—and then switch to intrusion detection at night, making the most of their investment.
*Testing was conducted with a 32-core HPE ProLiant DL320 Gen11 with three (3) NVIDIA L4 Tensor Core GPUs and 128 GB RAM. Number of supported camera streams may vary depending on the AI engine(s) being used.
Prepare for the future of retail
Retailers need a comprehensive AI solution that can analyze video data in real time, mine video streams for rich intelligence, and accelerate forensic video search. These capabilities are key to enabling faster decisions, threat detection, and store optimization.
The HPE hardware and software solution powered by NVIDIA and IronYun leverages advanced AI to offer intelligent video insights. The TD SYNNEX HPE Vision AI Discovery to Deployment process from proof of concept to site assessment absorbs complexity, with integration and post-sales support as an added value for this solution.
Check out the solution overview to learn more. Visit HPE at NVIDIA GTC, a global AI conference running March 18–21, online and at the San Jose Convention Center. Register now!
Let us help you prepare for the future of retail. Contact us today.
Bradley Sweeney joined HPE in August 2022 and now leads AI Solutions for the worldwide Compute Business Unit. His charter is to drive the development of emerging AI solutions on HPE ProLiant servers with a focus in vision AI, speech AI, and generative AI.
HPE_Alliances
HPE Alliance Partners