As more and more artificial intelligence (AI) powered security cameras are installed in homes, providing homeowners enhanced security features including timely intruder alerts or notifications when their loved ones arrive home, companies are fighting for a stake in this market—worth USD 8 billion worldwide in 2019 and expected to rise to USD 13 billion by 2023, according to research institution Strategy Analytics.
Wuhan-headquartered computer vision company SimShine, founded in 2017, is one of them. The company’s new line of home security cameras, named SimCam, adopts AI edge computing, a technology that supports cloud-free data storage and analysis.
In this burgeoning and competitive market, where established companies such as Amazon-owned Ring, Chinese electronics vendor Xiaomi, and Netgear’s Arlos are all strong players, SimShine wants to differentiate itself by offering AI security cameras with improved latency and better privacy protection.
The firm raised USD 8 million in its pre-Series A round last October, which will be used for research, development, hiring, and marketing for its sub-brand SimCam, in a round led by Cheetah Mobile, with participation from Skychee, Skyview Fund, and Oak Pacific Investment.
“In the long term, we want to become the ‘eyes of the smart home,'” said Joe Tham, the co-founder and the head of international sales at SimShine.He added that the future of smart homes will connect all kinds of AI-powered devices to create a living environment customized to the owner’s preferences, with SimCam playing a crucial role.
From computer vision to home security cameras
In 2015, Tham, a Malaysian Chinese who graduated from the National University of Singapore and a serial entrepreneur, co-founded a portable drone company with some of his Chinese friends, named Simtoo.
Two years later, in a new business venture, they established SimShine, an AI solution provider which initially mainly focused on computer vision technology, widely used in drone making. This tech is what allows drones to detect and avoid objects to ensure safe navigation.
The firm also started to work with industrial partners, exploring wider uses for computer vision technology. For example, the company is currently involved in an ongoing project with a client in Belo Horizonte, Brazil, embedding AI sensors and cameras in traffic lights to detect abnormal behaviors in a bid to increase public safety.
As the demand for smart home devices mounts, SimShine decided to explore consumer-level AI product development in 2018. After one year of research and production, including a Kickstarter crowdsourcing campaign, the company presented its first smart home security camera under the SimCam brand, during the 2019 Consumer Electronics Show (CES). Priced at just USD 139, the camera is able to use AI for facial recognition, location training, pet monitoring, and other functions, at a more affordable price than competitors. For example, Netgear’s spinoff firm Arlo offers a similar product at more than USD 200.
SimCam has managed to cut prices partly because of the co-founder’s experience in the hardware industry, according to Tham. During his entrepreneurial experience leading a Singapore-based firm developing a line of MP3 and MP4 music device players, when those devices were still common back in 2003, Tham cut prices by outsourcing production to factories in Shenzen, China’s manufacturing hub. The contacts he developed during that time have helped him to maintain low prices for the SimCam’s manufacturing.
However, apart from competitive prices, SimCam also hopes to distinguish itself from other competitors with advanced technology.
Less latency and more privacy
The company’s first home security device was able to be trained to monitor precise locations and objects, with the ability to distinguish between people, animals, and vehicles. In addition, the owner could create a database that allows the camera to recognize familiar faces, Tham explained.
The camera also supports Amazon, Alexa, and Google Assistant, enabling voice commands. The device saves all event-triggered recordings locally on a pre-installed microSD card, instead of transmitting the data to a cloud-based server for analysis, as most other AI cameras do. Edge computing, Tham explains, enables the camera to perform the data analysis in real-time, within the camera, or at the edge of the network, causing fewer latency issues and fewer delays in the transfer of data.
The firm’s new device, the SimCam S1, with a wider field of view and more functions, was unveiled just two months ago at this year’s CES in Las Vegas, together with its AI-enabled wireless doorbell Ango, which also performs AI processing onboard. The camera doesn’t need a cloud subscription to work, like other competitors, and maintains an aggressive price tag, at just USD 129.99.
“The AI technology we’re using leads to less latency and better privacy protection,” Tham said. “It’s because of the edge computing technology.”
As privacy remains a paramount concern for SimCam users, the use of edge computing, as opposed to cloud computing, reduces security risks, Tham says.
Cloud computing, the outsourcing of massive data storage and processing, is used widely in the internet of things (IoT) industry, including many smart home devices. Usually, cloud computing requires users to pay a monthly fee for the service which leverages remote cloud servers instead of buying expensive hardware. However, data stored in the cloud can be vulnerable to hacking and presents a higher security risk.
In comparison, edge computing is done at or near the source of the data, without relying on the cloud servers of different data centers hundreds of kilometers away. This increases speed and performance by reducing latency and protects the user’s privacy, as the encryption and storage of biometric information are done within the device. For companies, edge computing could also bring cost savings, as they won’t have to pay for bandwidth and cloud services.
The global edge computing market is expected to grow from USD 2.8 billion in 2019 to USD 9.0 billion by 2024, at a compound annual growth rate of 26.5%, according to report by research firm MarketsandMarkets, boosted by the increasing adoption of IoT across industries.
International ambitions
SimCam focuses its marketing strategy internationally, specifically targeting “more mature” markets including the US and Japan, Tham says. The company’s devices are sold through both offline and online sales channels, its official website, Amazon, Best Buy, and, Walmart, while also appearing on some Chinese e-commerce platforms, includingits own official website and Taobao’s Tmall.
“Consumers in different regions have different requirements for products,” Tham said. “Users in the States may care more about data protection and privacy, Japanese consumers have a higher demand for elderly-care, and Chinese consumers, as well as those in India, probably want more cost-efficiency.”
Last year, more than 20,000 units of SimCam’s products were sold worldwide. However, the company plans to sell 80,000 units this year.
On top of consumer-level product sales, SimShine also works with enterprise partners to provide AI-powered security solutions. For example, Tham said his company has started a project setting up its products at the entrance of a hospital in Southeast Asia to recognize visitors and patients. The company also plans to explore more scenarios, such as alerting nurses when a patient suffers a fall or gets into an emergency situation, although no other specific details where provided.
Amid the ongoing coronavirus outbreak, Tham said that both the company’s R&D center, located in Wuhan, and production lines, in Shenzen, have been affected. However, the loss is manageable due to forecasting and inventory management, he added.
For 2020, the firm already has new AI products lined up, such as a pet monitor and a car dash camera, which are scheduled to be released in the second half of this year.
This article is part of KrASIA’s “Inside China’s Startups” series, where the writers of KrASIA speak with founders of tech companies in the country