S&I Building Management Application

S&I is a leading building management service provided in Korea. S&I has deployed the Scenera PaaS to enable real time monitoring of events within S&I managed locations using cameras located in the buildings. The solution is capable of detecting events such as Intrusion, people falling, people fighting, analyzing X Ray images from bags for unauthorized devices, loitering and left behind items. Real time notifications of these events are sent to security personnel who can then respond to these events.

Visit Microsoft Customer Stories for more information on “Achieving Edge-to-Cloud Processing forDX with AI Capabilities in Facility Management”.

The Problem

Processing of camera fields is computationally intensive. Without Edge processing on the premises, the bandwidth requirements to deliver many live video feeds to the cloud becomes expensive. Alternatively, on Premise Edge computing can be applied to the live video feeds. This can be computationally intensive and require a significant deployment of hardware to handle all the camera feeds.

The question arises whether it is possible to distribute the processing load between the Cloud and the On Premises Edge so that a light processing is performed at the edge to limit the traffic from the Edge to the Cloud. The light processing algorithm is far less certain in its results than an intensive algorithm. This can be compensated for by using an intensive algorithm in the cloud which can filter out results generated by the light processing algorithm that are less certain.

The Solution

Scenera, TNM and Microsoft worked together to deploy an architecture that allowed the distributed processing of people counting. This entailed running constrained algorithms which are highly inaccurate on the edge in the customer’s premises. This enables either lower cost hardware to be used or a larger number of cameras to be processed on the edge processing hardware.

The results from the constrained algorithm have a large number of errors. To overcome this the results with a lower certainty are forwarded to a Network Edge Cluster with a more robust accurate algorithm. This algorithm is used to confirm whether the constrained algorithm has correctly detected a person or not.

The benefit of this approach is that it reduces the hardware costs on premises, reduces the bandwidth of video sent to the cloud while maintaining the highest possible accuracy of the results.

We deployed and measured the solution in the S&I facility. We measured the results of the solution by applying the solution to four camera feeds and counting people entering turnstiles. We measured the accuracy of the solution by monitoring the results generated by the camera and compared them to events that we were able to visibly inspect. Based on these measurements our accuracy was greater than 95% based on a relatively small sample of 100 images.

  • Compared to a solution that shipped the encoded video to the cloud, the Edge Video System’s (EVS) split architecture transferred 22X lower data to the cloud.
  • EVS’s on-prem edge usage was minimal and only CPU-based; no GPU was used on the on-prem edge. Also, for GPU usage, our cascaded pipeline reduced the need for the heavy AI model by two-thirds.
  • The overall latency of the solution was a few 100s’ of milliseconds.

The distributed architecture that was implemented demonstrated a significant reduction in network bandwidth, a reduction of edge hardware costs while maintaining a high degree of accuracy with low latency.

With all the initial testing and installations of the system at LGE’s buildings, S&I is ready to commercialize EVS in its building management application globally.

This article was published on Microsoft Customer Stories on November 1st, 2022.

Achieving Edge-to-Cloud Processing forDX with AI Capabilities in Facility Management

This article was published on Microsoft Customer Stories on November 1st, 2022.

The demand for pro-active digital transformation (DX) continues to increase across industries, even for businesses traditionally considered to be conservative or slower at adapting, such as facility management operations. S&I Corporation is a leading building management service provider that has deployed a system enabling real-time monitoring of events at their managed locations nationwide in Korea.

This innovative solution can detect events such as intrusion, accidental falls, altercations, loitering, and even security scanning and image analysis of baggage for unauthorized items and devices. As a previous affiliate of LG Group until recently, S&I has long been pursuing implementation of a comprehensive suite, such as this, to provide solutions for the multi-faceted needs of building management and operations.

S&I Corporation collaborated with key partners including TNM, Scenera and Microsoft to launch this innovative system, built with unique architecture to handle the complex levels of distributed processing, ultimately enabling Edge-to-Cloud. The S&I Building Management Application (BMA) is now considered one of their pillars for increased business expansion that takes advantage of Microsoft Cognitive Services (AI).

Hyoung WonJoon, CEO of S&I Corporation says, “S&I is undertaking a new strategic vision to transform the company in the next five to seven years. We expect our newly developed service suite to be central to our ongoing success and aim to continue expanding its features and capabilities.”

Combining On-Premises and Edge Processing for Improved Monitoring

A typical enterprise-level building can have dozens, or even hundreds, of security cameras just on their first floor. The physical monitoring required is significant in terms of staffing and any automated on-premises monitoring of these camera fields is computationally intensive.

Lee Jong Ho, S&I CTO says about the specific challenges, “Lack of edge processing on-premises means considerable costs to handle bandwidth requirements as live feeds increase in their number of connections to the cloud. At the same time, deciding for on-premises edge computing requires a corresponding deployment of hardware to handle all the camera output.”

Essentially, the most ideal level of light processing is limited relative to traffic capacity from Edge-to-Cloud̶—a light processing algorithm is far less accurate in its results than an intensive algorithm.

The scale of staffing required for on-site facility management just for security monitoring is quite substantial. S&I covers upwards of 1,000 large-scale buildings in the Korean market. The manpower necessary to maintain all the related systems alone numbers in the hundreds.

“Almost all our customers have installed CCTVs for monitoring so there was already demand for AI-supported detection, meaning that security guards would no longer be required to provide full-time monitoring. We were waiting for quite some time to have a dependable system that could handle the required scale and complexity,” says Hyoung Won Joon, S&I CEO.

Achieving Accuracy at Scale with Cascaded Pipelines for Edge-to-Cloud

Research into possible approaches and solutions started in the field of facility management several years ago. S&I, along with many competitors, was trying to develop their own IoT technology to achieve this. At that time, S&I was introduced to Scenera through one of the company’s local partners, TNM. Scenera saw huge success in syncing and connecting AI, IoT, and 5G, demonstrating the many possibilities of integration with Microsoft Azure Cloud.

The Korean government provides certain levels of financial support for new technologies which aided in fast-tracking milestones as S&I sought to enable digital transformation for the services offered to its clients. The CEO & Co-Founder of Scenera, David Lee, is well-known in the Korean electronics industry for playing an integral role in establishing the widely-adopted protocol for HDMI technology.

The partnership developed between S&I, Scenera and Microsoft quickly investigated solving related issues, such as how to utilize the cloud and local use of bridge devices most effectively. One key factor was whether to select a closed or open solution to manage edge-based bridge devices and computing in the cloud. It became apparent that the most optimal approach was the open architecture provided by Scenera and the protocols developed by the Network of Intelligent Camera Ecosystem (NICE) Alliance.

Heo Jai Ho, S&I Product Owner says, “We realized we would be able to better implement AI and other technologies in an open development system with Scenera. From there, we were well on our way to building a solution that we could actually use and sell to our demanding multinational customers.”

One of the notable distinctions when S&I developed its application was that it could leverage existing CCTV camera systems that were already in place. Scenera was instrumental in demonstrating how to encode, store, and stream video and audio at scale with Azure Media Services. The BMA platform also utilizes Azure Stack Edge as a cloud storage gateway and enables eyes-off data transfers from this existing infrastructure. That meant customers did not have to face the hurdle of removing their existing infrastructure to replace it with newer technology.

“Scenera’s platform allows our clients to go live immediately with a consistent and scalable architecture that takes advantage of existing hardware connected to the cloud and powered by AI. This was something no other competitor could do,” says Lee Jong Ho, S&I CTO.

Seamless Realization of AI-Powered Digital Transformation

As one of the major players in the building management industry in Korea, S&I typically serviced large-scale facilities for enterprises. However, an increasing number of opportunities are expected to develop with small and mid-sized building owners. The key points of contact vary greatly from office managers, building owners, and actual tenants, along with their respective needs and requirements.

The previous points of contact at larger facilities were office managers, but with S&I shifting its focus an approach is needed more suited to building owners and tenants at smaller locations. Self-design sessions for this new category of tenant are a large part of the consultation process and S&I is now able to include in that scoping the capabilities of its management application. At the same time, S&I is able to offer both scalable data services with Azure Storage, and a level of Enterprise security to these new SMB customers as the application integrates Azure Key Vault.

A typical contract for larger buildings usually includes 50 people for cleaning and security on a 24/7 basis for 3 years. In a mid-sized or smaller building, service components must be divided into independent items rather than being provided as an entire package. Therefore, the goal with these micro-service components is to develop or repackage them as digital options whenever possible.

An example of this is janitorial service which covers 50 cleaners in a building to clean floors, toilets, and offices. However, robots have now been developed that allow 1 cleaner to cover work previously done by 7 to 12 people and that service can be bundled into a single product. Through the BMA platform, cleaning services can be ordered and managed as needed throughout the week or weekend at both larger and smaller facilities.

Heo Jai Ho, S&I Product Owner elaborated on this new approach saying, “The purpose is to allow tenants in small and medium-sized buildings to easily log onto our management platform and request specific types of services as needed. This is a very different way of selling our business and the services we provide. Thus, when available as a platform, offerings that were previously available only to the large building segment are now easily accessible by small and medium-sized businesses.”

A Solid Foundation for Regional, National and Global Growth

S&I has been able to successfully map out and achieve a phase-based approach while re-positioning its service portfolio. Phase 1 was the research and implementation of an AI-based solution for facilities management which led to the partnerships with Scenera and Microsoft and large-scale implementation of the Azure Product Suite. Phase 2 was implementation and roll-out at the wide variety of facilities owned by LG Group.

S&I has a related application called Hi Pineapple, which is currently provided to 550 tenants, and it is becoming very popular. There are now approximately 60,000 monthly users making orders and requests, such as adjusting room temperature, whether they want a room cleaned, a meeting room reserved, and even what buses can be taken to specific destinations. The capabilities of this application make it easy to eventually adapt as a broader-type of consumer-focused software and service offering. Phase 3 will see the company’ s platform services such as this provided to a wider range of small, mid-sized and larger facilities.

Hyoung WonJoon, CEO of S&I emphasized the significance of this phased approach saying, “Our transformation is ongoing and we are aiming to build out our management platform not only for ourselves but also other suppliers, players, and partners as service providers. And this is on the demand side. Building owners who want better asset management can still choose from a menu of features they need. Our tenants with small and medium-sized buildings will have access to a full range of services at the push of a button, including security, cleaning, and facility management-related services.”

S&I intends to thoroughly address potential in the Korean market first and then turn its focus to expansion overseas. With their phased approach, technology and partnerships with Scenera and Microsoft, the executive team plans to become a 700-million-dollar company in terms of sales by the 2027.