918博天堂·(中国区)首页

Blog
Topic
All
AIoT
SMB Solutions
Products and technologies
Industries
All
Retail
Traffic
Education
Logistics
Banking
Building
Energy
Manufacturing
Sports
Sustainability
Business trends
Cybersecurity
Partner and customer experience
Alarm
Access control
Video Intercom
Traffic Management
LED
Intelligent Manufacturing
Resources and events
Search blog
Blog
Filter
Cancel
Topic
All
AIoT
SMB Solutions
Products and technologies
Industries
All
Retail
Traffic
Education
Logistics
Banking
Building
Energy
Manufacturing
Sports
Sustainability
Business trends
Cybersecurity
Partner and customer experience
Alarm
Access control
Video Intercom
Traffic Management
LED
Intelligent Manufacturing
Resources and events
Reset
Submit

Irida Labs and Hikvision: Pushing the boundaries of AI at the edge

AI innovator Irida Labs has integrated PerCV.ai, its software and services platform, into the Hikvision DeepinView AI camera series. Applied in the industrial and warehousing markets, the AI-enabled cameras help logistics companies and manufacturers monitor their operations in a fully automated way. Hikvision interviewed Demetris Anastassiou, Product Marketing Manager at Irida Labs, to find out more about the solution, and how it is pushing the boundaries of embedded AI to deliver new levels of operating efficiency for customers in multiple industries. 

 

Q: Could you please tell us a bit about Irida Labs? 

A: Irida Labs is an embedded AI software leader, with more than 10 years of experience in Computer Vision and AI at the Edge. Powered by a strong team of over 25 engineers, our work enables companies around the world to deploy scalable, vision-based solutions to monitor and streamline a range of processes in different verticals, from automation of goods tracking in logistics to monitoring of production processes in manufacturing and health & safety monitoring in any work setting. 

 

Q: Can you provide some background information about your flagship offering, PerCV.ai. What is the solution and how does it work?

A: PerCV.ai enables the end-to-end deployment of application-specific AI solutions that run at the edge, which means very near or even inside the cameras, with no server or cloud involved. It can be applied to many different vertical markets and use cases, but specifically for the industrial market it allows logistics companies and manufacturers to monitor their operations far more comprehensively and accurately, with no need to manually review video footage. To do this, it analyzes the camera stream in real time and creates analytics, along with point-in-time 'snapshots' of events that matter, such as a pallet or vehicle moving to a different location within a warehouse. These combined metadata can then be cross-referenced with data from existing systems, such QR readers or RFID readers, to create a comprehensive, accurate, and current view of processes, shipments, orders, and inventory. 

 

Q: What are the top benefits of PerCV.ai for customers? 

A: Our PerCV.ai-powered solutions open a whole new world of business insights and analytics. For the manufacturing or logistics industry, the top benefit is the ability to monitor processes and goods without manual effort, which significantly increases efficiency, reducing overall operating costs. In addition, the solution helps to ensure that goods are tracked through the supply chain from end to end, enabling unprecedented control over each step of the process. It can also be used to improve health & safety compliance with automated Personal Protective Equipment (PPE) monitoring – all in a fully automated way. New use cases for the technology are emerging all the time. 

 

Q: PerCV.ai uses 'edge processing'. Why is that so good for customers? 

A: Doing AI at the edge provides great cost and efficiency benefits. With all the AI analytics carried out in the camera itself, there is no need for centralized servers or expensive cloud resources. The fact that video streams are not transmitted means low bandwidth requirements and network costs, and the fact that out of thousands or hours of 24/7 footage only a few kb of useful metadata are kept supports >95% data storage compression rates. Also, the business data never leave the premises, and the video streams are not stored and cannot be retrieved, which results in better data privacy and safety.

 

Q: You use Hikvision cameras for the PerCV.ai solution – why is that? 

A: The Hikvision cameras are a perfect fit for a number of reasons. Most importantly, they offer a dedicated onboard processor for AI applications. This feature makes embedded optimization for our software far easier, saving both time and development costs, with no need for extra, third-party integrations. These cameras also give us all the computing and storage resources we need to deliver real-world, premium AI solutions to our customers. 

 

Q: Was it easy to embed the PerCV.ai solution into the Hikvision cameras? 

A: Yes, it was, thanks to Hikvision's HEOP platform. This is a development interface that allows us to deploy our software on the camera, ensuring that all the required processing and storage resources are available. Normally, embedded optimization of AI software into edge devices requires consideration of all kinds of parameters from the speed of the processor to the memory footprint. With HEOP, the requirements and limitations are much more transparent, which makes the integration process faster and easier. 

 

Q: How is Hikvision helping you to achieve your business and technology goals?

A: For AI to realize its market potential, it must work in real-world settings. With Hikvision, we have been able to integrate our software with the Hikvision DeepinView AI camera series, and to create just such a solution. It means that the barrier of entry for AI is significantly reduced, which is great news for our customers, and for the industry as a whole.

Q: What's the best thing about your partnership with Hikvision?

A: First, the Hikvision cameras offer the processing speed and capacity needed to support PerCV.ai, and the HEOP platform made it fast and easy to embed our solution into the camera. Hikvision is also a household name, and many customers also have experience of working with Hikvision cameras, so it's a familiar technology that will smooth customers' journey to AI.

text over image1754
  • Hikvision cameras offer the processing speed and capacity needed to support PerCV.ai, and the HEOP platform made it fast and easy to embed our solution into the camera.

  • Hikvision cameras offer the processing speed and capacity needed to support PerCV.ai, and the HEOP platform made it fast and easy to embed our solution into the camera.

Other than the technical aspects of the partnership, we've benefitted hugely from close and constant collaboration with the Hikvision team. Until recently, we were having weekly meetings with them, and we've always had help to resolve any technical questions or issues that have come up. It's been a great, supportive relationship that has delivered great results for our business and our customers.

 

Q: What are your future plans for partnering with Hikvision? 

A: Because of our close working relationship with the Hikvision R&D team in Europe, we expect to continue pushing the boundaries of what's possible with Hikvision cameras. The team has already helped us to adjust specifications to support new features in our solution, and to update the device firmware, and we are also working together to develop AI and automation solutions for new vertical markets. All this means we can continue to expand the AI market and achieve even better outcomes for our customers. 

 

Find out more

If you'd like to find out more about the Irida Labs PerCV.ai solution and the Hikvision HEOP platform, please contact us today.  

 

Related links

Subscribe to newsletter

Subscribe to our email newsletter to get the latest, trending content from Hikvision

Contact Us
Hik-Partner Pro close
Hik-Partner Pro
Hik-Partner Pro
Scan and download the app
Download
Hik-Partner Pro
Hik-Partner Pro

Get a better browsing experience

You are using a web browser we don’t support. Please try one of the following options to have a better experience of our web content.