Arcules: Providing 4x faster video cloud IoT service

About Arcules

Arcules, a Google Cloud partner, transforms video into powerful business intelligence by aggregating video and IoT data using AI and cloud technologies. The company’s solutions help enterprises optimize their business operations and make the world a safer place.

Industries: Technology
Location: United States

Arcules leverages the Google Cloud Platform to provide its customers with a unified solution that analyzes data from video cameras and IoT devices for actionable insights that improve security measures and boost top line growth and operational efficiencies.

Google Cloud Results

  • Boosts performance with a highly secure cloud solution that operates as fast as on-premises video systems
  • Provides a unified platform for analyzing IoT and video data, enabling customers to make actionable insights to increase revenue and reduce costs
  • Moves traditional security practices from reactive to proactive and preventative

Writes and retrieves data from Google Cloud Bigtable in microseconds

Arcules, a Google Cloud partner, is a fast-growing Canon (NYSE: CAJ) company spun out from Milestone Systems. It was launched to pursue new opportunities based on Milestone’s 20 years in on-premises video monitoring and video management software solutions. Announced in 2017, Arcules takes a cloud services approach to synchronizing and mining data, compiled from video cameras and IoT devices, for clients who want insights to improve safety, operational efficiency, and top line growth.

The Arcules platform enables more robust insights for retail foot path, sentiment, and dwell-time analyses, using video cameras and IoT devices to correlate data on shopper behaviors in store. Cross-sensor data aggregation and correlation allow brands and retailers alike to obtain information previously used only for surveillance purposes.

For customers concerned about security threats on their premises, the Arcules solution leverages artificial intelligence and machine learning to help automatically detect, in real time, if a person appearing on a video feed might be in an area they should not be in.

“The performance difference with Google Cloud Platform was amazing—superior to the others by a factor of four. Combined with security features in Google and better pricing, it was clearly the best strategic technology choice.”

Andreas Pettersson, CEO, Arcules

Managed services let Arcules focus on innovation

As it evaluated a number of cloud service providers, Arcules understood latency would be a key challenge due to the high bandwidth of video data. During testing and development of its solution, the Arcules IT team expected to see up to 8 seconds of latency. While other cloud providers demonstrated significant latency during a competitive evaluation period, Google Kubernetes Engine on Google Cloud Platform (GCP) showed just 1.2 seconds of egress latency for live video right out of the gate.

Storage was also a major factor for the company, since this is a platform that will handle a lot of video and IoT data. With the wide range of GCP storage options and guaranteed IOPS, this helped Arcules find the best option for this data to be analyzed by its platform.

“The performance difference with Google Cloud Platform was amazing—superior to the others by a factor of four. The others couldn’t get close. Combined with security features in Google and better pricing, it was clearly the best strategic technology choice,” says Andreas Pettersson, CEO of Arcules.

Eliminating mundane infrastructure management tasks such as security patches and upgrades helps free up the Arcules team to focus on higher-value, customer-centric solutions. “For a company like Arcules, through our startup phase and our preparations to scale massively, very quickly, the managed services provided by GCP fit the bill perfectly. As a result, we can focus on innovative ways to add value for our customers,” Andreas says.

With the company’s Google Cloud infrastructure in place, it easier for Arcules to remotely handle firmware updates for devices on its customers’ premises, quickly and with minimal expense. Previously, some of its customers struggled with the on-site video devices and servers used by other video monitoring companies. They were subject to downtime or lengthy maintenance cycles, often requiring months of wait time for security patches and costly visits to the sites by technicians.

“The up-to-date security of our cloud-based solution, plus all the intelligence the Arcules service offers, are massive benefits that legacy video monitoring systems can’t provide. Google Cloud Platform helps make it all possible.”

Andreas Pettersson, CEO, Arcules

Better security in the cloud with GCP

In legacy video monitoring systems, any open ports could be subject to exploitation if firewalls are not properly maintained. Arcules recognized the security advantages offered by GCP for bolstering the end-to-end security capabilities of its own service offering.

NCC Group, a Google Cloud Security Services Partner, set up the GCP policies and permissions, in addition to conducting threat penetration testing to help Arcules strengthen its systems.

“The up-to-date security of our cloud-based solution, plus all the intelligence the Arcules service offers, are massive benefits that legacy video monitoring systems can’t provide customers. Google Cloud Platform helps make it all possible,” says Andreas.

Marrying video and IoT data for business insights

Data from video cameras and other IoT devices goes into the Arcules gateway, an on-premises Kubernetes cluster that pushes data to the Arcules Intelligent Cloud Video platform, running on GCP. With high volumes of video data a key consideration, Arcules puts the video data into buckets and pushes it quickly from the gateway to the cloud. Customers ultimately benefit as the hardware needed on site is minimized thus reducing capital and operating expenses.

Arcules is a heavy user of Google Kubernetes Engine. “Leveraging microservices via Google Kubernetes Engine allows us to scale up and down based on demand at any given time,” says Sheriff Mohamed, Principal Engineer for Arcules. “Further, it enables Arcules to spin up engineering environments with minimal disparity.”

“Since we’re using Google Kubernetes Engine, we don’t have to worry about the master node. That’s an extreme plus for us, because it would be a very risky piece of the system for us to have to manage on our own,” adds Sheriff.

Operations and management are simplified with use of containers and Google Kubernetes Engine. “Because the environments we create are running on Google Kubernetes Engine, they’re all nearly identical other than the scale. So, if we do performance testing in one environment, we can be confident that it’s going to be similar to when it goes into production. Everything becomes a lot more uniform, driving efficiency for our operations team,” says Sheriff.

Training neural networks for better intelligence

Arcules uses Google Cloud Pub/Sub as the message bus in between its microservices, for superior scalability. The microservices connect the neural networks the company is retraining, using TensorFlow, based on the customer’s desired parameters.

Most of the unstructured video data becomes structured once it is pushed into GCP and married with data from IoT devices and other sensors. “The data from devices is very important to us, because it often reveals the relevancy of the video data,” says Sheriff.

“Our latency in writing the data into Google Cloud Bigtable and extracting it is extremely fast—literally microseconds. It’s much faster than we had anticipated going into this and it’s exactly what we need.”

Sheriff Mohamed, Principal Engineer, Arcules

For example, a motion sensor by a retail store’s featured promotional display can help detect activity and be indexed against video data of a shopper’s path through the store. The neural network can be programmed with TensorFlow to categorize and learn what actions a shopper takes after viewing the featured display, triggering notifications and insights. Did they pick up the item? What was their sentiment? Did they consult their phone? What other items did they pick up? Did they make a purchase? With analysis of this type of information, the store can make adjustments to influence shopper behavior.

Time stamping is critical for marrying data captured across devices in a store. “We use Google Cloud Bigtable for video sequence data and with some of our other sensors, primarily for its time series capabilities. By taking the time series data in directly, our latency in writing the data into Cloud Bigtable and extracting it is extremely fast—literally microseconds. It’s much faster than we had anticipated, and it’s exactly what we need,” says Sheriff. The speedy results, leveraging large amounts of data, impress the Arcules team and its customers alike.

In order to catalogue its growing roster of clients, Arcules uses the fully managed MySQL and PostGres database services provided by Google Cloud SQL. This handles customer database information on permissions, subscriptions, authentication, and other customer profile data.

“The Google Cloud Platform overall is amazing. We’re going through many different virtualized layers of network, yet the speeds we see are substantial. High performance is a must-have in order to quickly deliver actionable insights for our customers,” says Sheriff.

A next-generation video analysis system

In the near future, Arcules would like to foster a developer community using its APIs to create solutions on top of its core Video Cloud IoT offering, which it expects will also leverage Google Cloud Platform.

“We are changing an industry with our unified solution for analyzing data from video cameras, IoT devices, and more. Our customers are very excited about the robust business intelligence, not previously available, that they can now leverage to improve operations and drive competitive advantage,” says Andreas.

About Arcules

Arcules, a Google Cloud partner, transforms video into powerful business intelligence by aggregating video and IoT data using AI and cloud technologies. The company’s solutions help enterprises optimize their business operations and make the world a safer place.

Industries: Technology
Location: United States
Google Cloud Platform logo

12 Months FREE TRIAL

Try Kubernetes Engine, BigQuery, and other Cloud Platform products with $300 in free credit and 12 months.

TRY IT FREE
Google Cloud Platform logo

12 Months FREE TRIAL

Try Kubernetes Engine, BigQuery, and other Cloud Platform products with $300 in free credit and 12 months.

TRY IT FREE