California-based SiMa.ai is collaborating with Lanner and AWL to accelerate smart retail at the edge by combining hardware with machine-learning software and video analytics.

SiMa.ai is a software centric, embedded edge machine-learning system-on-chip company. It is working with Lanner, a Taiwanese provider of industrial computing appliances, and Japanese AI software company AWL, a specialist in video analytics. The partnership aims to deliver high-performance AI at the edge for manufacturing, retail and transportation.

The integration combines Lanner’s EAI-I730 and LEC-2290 edge AI appliances with the SiMa.ai machine-learning system-on-chip (MLSoC) PCIe card platform and Palette Edgematic software, along with AWL’s AI-based video analytics software.

This should let users rapidly deploy AI inferencing at the edge, providing real-time multi camera video analytics to interpret ingested frames on the fly. Each SiMa PCIe card delivers up to 50Tops performance. The LEC-2290 takes one PCIe card whereas the EAI-1730 can incorporate up to four PCIe cards.

“SiMa.ai Palette software and MLSoC empowers customers across multiple verticals, such as manufacturing, retail, aerospace, defence and healthcare, to deploy high performance AI applications at the edge,” said Elizabeth Samara Rubio, chief business officer at SiMa.ai. “The partnership with Lanner and AWL allows us to deliver a joint AI offering making it easier for customers to scale embedded AI applications at the edge for retail and manufacturing use cases.”

Maulik Upala, director of AI at Lanner Electronics, added: “SiMa.ai is the clear leader in bringing computer vision intelligence to any device in the edge market through its remarkable fusion of hardware and software innovation. In supporting the industry’s widest range of ML models and ML frameworks for computer vision, SiMa.ai ensures our customers can focus on rapid innovation versus implementation.”

And Yasuhiro Tsuchida, CTO of AWL, said: “Our collaboration with SiMa.ai enables our customers to deploy real-time, high-performance, accurate retail analytics, enabling them to simultaneously reduce both cost and power-consumption. The ability to accelerate complete workload pipelines in the SiMa MLSoC, as opposed to ML acceleration only, enables us to provide differentiated, highly performant optimisations for retail environments that are sensitive to cost and power overheads.”

SiMa.ai delivers one platform for all edge AI that scales with users as their AI and ML journey evolves, from computer vision to transformers to multimodal generative AI. The SiMa.ai MLSoC enables full pipeline deployment of complete real-world workloads as a standalone edge-based system.

The MLSoC works seamlessly with Palette software to empower users across multiple verticals such as industrial manufacturing, retail, aerospace, defence, agriculture and healthcare, with increased compute capabilities, while increasing efficiency by delivering high frames per second per watt (FPS/W) performance in the edge AI and ML market.

SiMa.ai (sima.ai) was founded in 2018, has raised $270m and is backed by Fidelity Management & Research, Maverick Capital, Point72, MSD Partners, VentureTech Alliance and more.

Lanner (www.lannerinc.com) is a provider of network appliances, delivering technologies for telecommunications, enterprise and industrial applications.

AWL (awl.co.jp) is a certified start-up from Hokkaido University that has a number of achievements in using and introducing IoT, mainly in the smart retail industry, using AI video analysis technology. Its proprietary AI technology supplements the human eye and visualises people and real spaces in real time at any site.