Delivering Peace of Mind in IoT - Using a Reliable Cloud Framework for Services
Head Division for Global Development - Solutions Development Division - IoT Camera Systems Development Section
Pictured: Town Recorder group
TOA Corporation is a company based in Kobe that specializes in audio and video recording devices. TOA develops, produces and sells industrial-use devices under a BtoB model, and has a global reach.
This case study is an example of a company’s transition of production to a full-managed, cloud-native environment centered around PaaS (Google App Engine). We (Cloud-Ace) had the honor of having TOA speak during the main session at our OPEN DX 2020, where they presented this case as a realistic scenario of Google Cloud integration for production businesses in West Japan. We hope the reader can use this case study to understand how problems were solved, and what effects they had.
Town Recorder Remote Monitoring Service Overview
Town Recorder is an integrated system of a FHD camera, recording storage device, and LTE/LPWA/Wi-Fi capabilities which results in an easy to install, all-in-one camera that works as long as there is electricity.
It is being utilized for surveillance in areas such as shopping districts, school zones and intersections, making these areas safer for the local community. Perhaps you have seen one of these devices on utility poles or lampposts in the city.
In addition, the Town Recorder Remote Monitoring Service enables users to remotely check the operational status, and even control, the Town Recorders. As Town Recorders include Dead Spot Surveillance function, Live View function, External Devices Limitation function, Long-Range Broadcast function, remote download of recordings, and abnormal situation emails, they can be utilized in diverse circumstances beyond crime and fire prevention.
Challenges in the Service
In light of replacing the system, 3 objectives were to be achieved.
First, reduction of cost for maintenance, and reduction of cost and labor during changes in deployment scale. Our old system had multiple virtual laaS servers, and we wanted a unified managed PaaS system for our new solution.
Second, a consistent running cost, regardless of amount of usage. Our old laaS system was a flat monthly payment plan, and by switching to a pay-per-use plan we aimed to optimize our running costs.
Third, increasing customer confidence. In order to make the system available to view at any time, especially during emergencies, we required a cloud system with high reliability, so we decided on using Google Cloud.
Background for the Employment of Google Cloud
In considering the foundation for our next cloud system, we put special consideration into the reliability and possibilities of the infrastructure. Since Google™ is self-sustaining in everything from network cables and servers to datacenter management as well as being able to cater for everything from hardware to software, we were intrigued by their offering.
The availability of TPU on Google Cloud fit our consideration of future AI integration, and their thought to the environment through usage of renewable energy and energy saving measures convinced us to go for Google Cloud. We also concluded that the cost-performance ratio offered by them was high compared to other cloud service providers.
Why We Chose Cloud-Ace
The catalyst was our visit to Cloud-Ace’s booth during a 2018 exhibition. We became interested in Cloud-Ace and Google Cloud, so asked to speak with them further. Our interest increased during our discussion when we found out that Cloud-Ace is a leading company as a Google Cloud vendor, and that they are also a Google Cloud Premium Partner. Above all, we were especially drawn by the impressive speed and tempo of our conversation on development plans.
For the “information pipeline” of the cloud, there is the “camera to Google Cloud” pipeline and “PC to Google Cloud” pipeline. For the camera pipeline we utilized Cloud Functions, and for the PC pipeline we utilized App Engine. This allows us to make flexible decisions when there is a change in the number of cameras, or when we need to change operational scale due to change in number of viewers, or during an eventual expansion of functionality.
Since our old system was entirely structured on laaS, we used to have to constantly monitor the processing pressure caused by data inflow from the camera and PC pipelines. We also had to constantly change up the number of operational servers, along with the mundane maintenance tasks entailed by this process. We feel that the decrease in required maintenance made possible by our system overhaul is a big benefit.
In addition, pertaining to the organization of logs produced by our devices, keeping in mind the fact that camera systems in business scenarios tend to be operated for a longer period of time, there is a need to keep an accurate track of both current and past logs. As an increase in cameras means an increase in the number of logs, we decided to employ Cloud Bigtable™ after considering cost-merit factors.
For system login, we saved development time and increased security by implementing Firebase™. We also utilized a few other products of Google Cloud and other companies in our overall system.
Advantages and Effects of implementing Google Cloud
By implementing PaaS across the board, we were able to decrease operational costs of infrastructure. In addition, we were able to decrease downtime caused by system changes and expansions for our existing users, thus increasing the quality of services we provide. For the actual everyday operations, Google Cloud’s easy-to-use interface allowed us to make the development process smooth. We especially made use of the SQL results export, log display and app update history functions.
We also mainly used G Suite™ (Google Workspace since October 2020) to communicate and exchange information with Cloud-Ace. Everything including account management was extremely convenient since the service was built on Google Cloud, and our productivity went up.
Since we want to make use of AI in the future, we look forward to making use of Google Cloud’s diverse and cost-effective AI products.
In addition, we want to implement audio and video sensing capabilities into peripheral devices apart from the external sensors, integrate with other business’ services through the cloud, and contribute to the development of cities’ infrastructure. Finally, we want Town Recorders to provide safety and respite to people, increasing quality of life and productivity.