Hawk Hsieh, SpotCam: “AI will have the opportunity to improve your life”
As technology advances, we are finding new ways to examine our surroundings. Newly introduced cameras have the capacity to detect human activity, monitor humidity and temperature, and perform sophisticated face recognition. But how safe are they?
Cameras are often associated with security and protection. But they can also be used for surveillance and other malicious purposes. So where is the borderline between “useful” and “dangerous,” what makes a camera secure, and what is the future of AI in video platforms?
Let’s go back to the very beginning of SpotCam. What has the journey been like over the years?
It’s not easy, we are one of the early suppliers that integrate cloud technology and IP camera, and mobile and web software to create an easy-to-use video surveillance solution for households and SMBs (small and medium businesses) since 2014. There are so many me-too products coming out afterward, and we focus on making our system more stable and creating more and more thoughtful features on our cloud and mobile/web client software so we can still stand out while so many brands shut down or leave this market over the years. The focus of the SpotCam solution can be categorized into three points: collect video and environmental data, show the data and allow users to make use of the data in the easiest way, develop a subscription business model based on value-added features on the cloud.
- Video and environmental data collection: during past years, various cameras are introduced to meet the surveillance needs of residential and commercial use. Such as cameras that not only collect video feeds but also collect humidity and temperature data of the environment; or cameras that are smart enough to detect if there is human activity in the scene.
- Display of monitoring information: we try our best to make it easy for users to obtain monitoring information, including video and other environmental data, on whatever devices they like to use, such as an Android mobile phone, iOS mobile phone, Google Home, or Amazon Alexa devices, web browser, IFTTT, etc.
- Subscription business model (SAAS): when all the monitoring data converges on the cloud, value-added to the monitoring service becomes possible. Extended video storage is one of the most basic optional services that can be provided. Users no longer need to worry about the high cost of storage equipment and bulky space. In addition, a real-time image AI algorithm is also applied so we can derive intelligence out of tons of video feeds and provide insight to our users, so they know what is going on.
Can you introduce us to what you do? How is AI incorporated into your products?
I am the CTO of SpotCam and responsible for cloud system architecture design and the development and maintenance (DevOps) of the SpotCam cloud platform. The platform scope includes real-time video streaming, video recording, and callback, real-time event push, cloud computing AI algorithm, and integration.
SpotCam deploys both edge computing and cloud computing AI capabilities. Edge computing is built into the camera, and you can get such AI functions when you buy a camera, while cloud computing AI requires a subscription and runs on the cloud platform we build. For the integration of cloud computing, we have deployed cloud computing hosts in every major region, including EU/US/ASEAN, and use streaming video inference for better performance. Therefore, we can provide real-time AI event pushes to users who subscribe to our cloud services.
Which myths and misconceptions surrounding smart home devices do you notice most often? What do you think they are based on?
The most common misunderstanding about smart home devices is that as long as the device has Wi-Fi capability and allows app remote control, then it is a smart home device. It is often advertised that people are going to remotely control the air conditioner at home or hold the mobile phone in bed to control the electric lights before sleep, which is really not useful and more like a fancy feature instead of a real need. This situation has been improved in recent years with the use of AI speech recognition. Conditional control can be done by speaking, for example: turning on the air conditioner when the temperature is higher than 28 degrees, turning on the light when I wake up, etc. In my opinion, the smart home devices must form an ecosystem, and the devices can intelligently act depending on different conditions and user scenarios. Then it is a real smart home device.
How did the recent global events affect your field of work? Did you add any new features as a result?
There is no doubt that the COVID 19 pandemic made many people work from home, including me. Especially parents needed to take care of their babies at home because baby caring centers are shut down. Therefore, we once again used cloud technology to integrate software, hardware, and cloud algorithms to make an innovative baby monitoring camera – SpotCam BabyCam, that is smart enough to detect baby crying sound and face cover or rollover during baby sleep, also making baby albums automatically to create happy family memories.
Besides video surveillance systems, what other security measures are essential for any critical infrastructure company?
In terms of safety, we have tried to use a video analysis approach to detect people's falls or abnormal behavior in operation sites of different mission-critical environments such as mining, oil, and gas, rolling stock, etc. It’s not enough to simply rely on video-based safety measures for the safety of critical infrastructure. It’s imperative to implement other special-purpose sensors or detectors, such as sensors that can detect abnormal vital signs of the living body without contact, for example, falls, apnea, abnormal heartbeat. Perhaps in the future, bio-radar can work with video surveillance to more accurately detect life-threatening events.
Talking about average individuals, what tips would you give to those looking to secure their homes and networks?
For sure, we are advocates for cloud-based home monitoring cameras. For average individuals, what you want is an easy-to-use surveillance system that doesn’t require the hustle and bustle for cabling and wiring, network configuration, storage equipment procurements, dedicated server spaces, etc. Cloud-based home monitoring cameras such as SpotCam products are designed for ordinary people who may not be tech-savvy and possess limited installation DIY capabilities. Another consideration is looking for brands that you can actually find references for on the internet instead of some E-commerce only brands.
There are many so-called importer/exporter brands that can only be found on e-commerce sites and nothing else. You can’t even find legitimate company information behind it. For that kind of product, you should worry about security and reliability, and the best way for general individuals to avoid any network security and privacy problem is to buy products from a brand that you can actually find legit references for on the internet.
What predictions do you have for the future of technology in the public safety field?
At present, a large number of CCTVs have been laid on the streets where people come and go. The purpose is to provide police units to quickly clarify the incident. In the future, 5G wireless communications with higher bandwidth and lower latency may bring new possibilities to video surveillance systems by making real-time ultra-high resolution video transmission a reality. Cameras deployed on a 5G network do not require complicated wiring and can be installed wherever power is provided. This combined with cloud storage such as SpotCam’s system, can make the video surveillance systems portable and easy to deploy. On top of that, with more and more AI algorithms added to the cloud platform, this portable and flexible video surveillance system will be able to execute more advanced and sophisticated jobs other than simply streaming live video feeds or storing the video data. We also believe in the future, what limits the functionality of a smart video surveillance system is not technology, but privacy and political concerns, such as face recognition, are being prohibited in public spaces now across the EU.
What other aspects of our lives do you hope to see enhanced by technology in the near future?
Now, AI has literally found its place and quietly improves our lives. The most obvious example is the license plate recognition system. The three elements of AI can be perfectly met in the parking lot scenario:
1. Data: Tens of thousands of cars enter and exit the parking lot every day, providing a stable source of license plate image data
2. Model: The labeling of the license plate is very clear. After training the model manually, it does not take long to train an AI model suitable for a parking lot.
3. Hardware: It’s more than standard to install cameras and lights in the parking lot and the camera video feed can be easily connected to the field control room where the computer server is located.
In the future, for those application scenarios that can meet the above three elements, AI will have the opportunity to improve your life. That’s also what we are seeking in SpotCam.
Would you like to share what’s next for SpotCam?
SpotCam will continue to introduce new hardware, including cameras and other security devices, all connected by our smart cloud platform to provide easy-to-use video surveillance and security service alternatives to residential and small business users. We will also further develop SaaS (Software as a Service) and VSaaS (Video Surveillance as a Service) business model by adding more and more value-added functions to our cloud platform so different users with different needs can subscribe to what they want and find their tailored video surveillance and security solution in SpotCam.