Skip to main content
Monthly Archives

March 2022

IoT Enabled Edge Infusion

By Blog

By Shikhar Kwatra, Utpal Mangla, Luca Marchi

The sheer volume of information being processed and generated by Internet of things enabled sensors has been burgeoning tremendously in the past few years. As Moore’s law is still valid providing an indication of number of transistors in an integrated circuit doubling roughly every two years, growing ability of edge devices to handle vast amounts of data with complex designs is also expanding on a large scale. 

Most current IT architectures are unable to compete with the growing data volume and maintain the processing power to understand, analyze, process and generate outcomes from the data in a quick and transparent fashion. That is where IoT solutions meets Edge Computing. 

Transmitting the data to a centralized location across multiple hoops, for instance, from the remote sensors to the cloud involves additional transport delays, security concerns and processing latency. The ability of Edge architectures to move the necessary centralized framework directly to the edge in a distributed fashion in order to make the data to be processed at the originating source has been a great improvement in the IoT-Edge industry.

As IoT projects scale with Edge computing, the ability to efficiently deploy IoT projects, reduce security to the IoT network, and adding of complex processing with inclusion of machine learning and deep learning frameworks have been made possible to the IoT network. 

International Data Corporation’s (IDC) Worldwide Edge Spending Guide estimates that spending on edge computing will reach $24 billion in 2021 in Europe. It expects spending to continue to experience solid growth through 2025, driven by its role in bringing computing resources closer to where the data is created, dramatically reducing time to value, and enabling business processes, decisions, and intelligence outside of the core IT environment. [1]

Sixty percent of organizations surveyed in a recent study conducted by RightScale agree that the holy grail of cost-saving hides in cloud computing initiatives. Edge AI, in contrast, eliminates the exorbitant expenses incurred on the AI or machine learning processes carried out on cloud-based data centers. [2]

Edge is now widely being applied in a cross-engagement matrix across various sectors:

  1. Industrial – IoT devices and sensors where data loss is unacceptable
  2. Telecommunication – Virtual and Software Defined solutions involving ease of virtualization, network security, network management etc.
  3. Network Cloud – Hybrid and Multi-cloud frameworks with peering endpoint connections, provisioning of public and private services requiring uptime, security and scalability
  4. Consumer & Retail framework – Inclusion of Personal Engagement devices, for instance, wearables, speakers, AR/VR, TV, radios, smart phones etc. wherein data security is essential for sensitive information handling 

Since not all the edge devices are powerful or capable of handling complex processing tasks while maintaining an effective security software, such tasks are usually pushed to the edge server or gateway. In that sense, the gateway becomes the communications hub performing critical network functions inclusive of but not limited to data accumulation, sensor data processing, sensor protocol translation and understanding prior to the gateway forwarding the data to the on-premise network or cloud. 

From Autonomous vehicles running different Edge AI solutions to actuators with WIFI or Zigbee with the ability to connect to the network, with increased complexity of fusion of different protocols, different edge capabilities will continue to surface in coming years.





Overview of Akraino R5’s Project Cassini: IoT and Infrastructure Edge Blueprint Family

By Akraino, Blog

By Olivier Bernard and Parag Beeraka 

This post provides an overview of the Smart Cities Akraino blueprint as well as an overview of key features and implementations of PARSEC in Akraino’s Release 5.

 Overview – Akraino Blueprint:  Smart Cities

The purpose of the Smart Cities blueprint family  is to provide an edge computing platform based on Arm SoC and  improve deployment flexibility and security within the edge computing. High-level relationships between  functional domains is shown in the figure below:

(For the full description of the Smart Cities Reference Architecture please refer to the Smart Cities Documents.)


Smart Cities in Akraino R5– Key features and implementations in Akraino Release 5:

 The Smart Cities blueprint’s security components is PARSEC, first available in Akraino’s Release 5. . The following is a brief introduction to PARSEC.

Parsec is the Platform AbstRaction for SECurity, a new open-source initiative to provide a common API to secure services in a platform-agnostic way.

PARSEC aims to define a universal software standard for interacting with secure object storage and cryptography services, creating a common way to interface with functions that would traditionally have been accessed by more specialised APIs. Parsec establishes an ecosystem of developer-friendly libraries in a variety of popular programming languages. Each library is designed to be highly ergonomic and simple to consume. This growing ecosystem will put secure facilities at the fingertips of developers across a broad range of use cases in infrastructure computing, edge computing and the secure Internet of Things.



(For more information of PARSEC:


Software Defined Camera (SDC) blueprint

Security Cameras are seeing increasing growth in many subsegments such as commercial surveillance cameras, consumer surveillance cameras and in many other devices such as dashcams, baby monitors and other IoT vision devices. The total surveillance market is expected to be around $44B in the year 2025 growing at a CAGR of 13%. The other exciting thing happening in this market, along with the increase in the unit shipments of surveillance cameras, is the increasing use of artificial intelligence and machine learning (AI/ML) in these cameras and IoT devices. It is estimated that there are already a billion cameras installed in the world and this number would reach 2 billion by the year 2024. However security cameras are very different from many other devices because once installed, they stay installed for more than 5+ years. So it is critical to ensure that these devices continue to provide enhanced functionality over time. 

In today’s world, there are many technologies that are “Software defined”. This has been possible because of many advancements in technology – at the bottom of the stack, you have the Operating System, then the virtualization layer – which actually kicked off the original software defined compute with Virtual machines etc. Then came containers and then the orchestrators for the containers such as k8s, k3s  – these technological advancements paved the path to define a software defined data center. In the datacenter, after software defined compute, software defined networking and software defined storage started to become the norm. 

Now the time is prime to move software defined trends to the edge devices as well. This is really a transformation that has started across many segments such as Automotive – cars, trucks etc. We do see this trend applied to many other edge devices and one of them would be cameras or as we and many others call it, Smart Cameras. The idea here is that once you buy a camera, the hardware is advanced so that you will be receiving continuous software updates which can be neural network model updates catered to the specific data that you might have captured using the camera or other updates related to functionality and security of the device.  


 By designing future camera products with cloud native capabilities in mind, one will be able to scale camera and vision products to unprecedented levels. If all the applications are deployed using a service oriented architecture with containers, a device can be continuously updated. One of the key advantages of this architecture would be to enable new use case scenarios that one might not have envisioned in the past with a simple on demand service deployment post sale through an app store. A simple example can be to deploy a new and updated license plate recognition app a year after the purchase of cameras. At the same time one of the other key advantages with this architecture would be to enable continuous machine learning model updates based on the data pertinent to the camera installations i.e be able to use the data that it is capturing and use it to train the models. With this, the model would be relevant to the specific use-case thus increasing the value of the camera itself. All of this is possible due to enabling a service oriented architecture model using containers and thinking cloud native first. 

The other important aspect that will really bring together the security camera and vision products is the adherence to common standards at the silicon and the platform level. LF Edge member company, Arm,  has started an initiative called Project Cassini and Arm SystemReady® , and it’s a key component of this initiative. As part of Arm SystemReady®, Arm is working with many ecosystem partners (OEMs, ODMs, silicon vendors) to drive a standards approach in order to scale deployments by replacing custom solutions with standards-based solutions and becoming the platform of choice for future camera deployments.  

To address and simplify these challenges, Arm is designing and offering the solution as an open source reference solution to be leveraged collaboratively for future software defined cameras based on established standards focused on security, ML, imaging and cloud-native. The basis of the reference solution is the SystemReady® certified platform. SystemReady® platforms implement UEFI, U-Boot and Trusted standard based Firmware. 

  • With the base software layers compliant to SystemReady®, standard Linux distros such as Fedora, OpenSUSE distros or Yocto can be run on these platforms. 
  • Then the container orchestration service layer can run containerized applications, by using k3S. 
  • The next software layer is ML and Computer vision libraries. Arm has supported ML functions through Arm-NN and the Arm Compute library so that users can develop their ML applications using standard open-source ML frameworks such as PyTorch, Tensorflow etc. 
  • Similarly, with computer vision, it is critical to support open standards such as Video 4 Linux, OpenCV and BLAS libraries. 
  • Both complementary and necessary, is security – this is where orchestration and security microservices such as PARSEC, an open-source platform abstraction layer for security along with PSA certification process are critical, so that end users build the confidence that the product will be secure when deployed in the field. These secure services are the foundation to support and enable containerized applications for inferencing, analytics and storage.

Learn more about Project CASSINI here:

Infusion of Machine Learning Operations with Internet of Things

By Blog

By: Shikhar Kwatra, Utpal Mangla, Luca Marchi

With the advancement in deep tech, the operationalization of machine learning and deep learning models has been burgeoning in the space of machine learning. In a typical scenario within the organization involving machine learning or deep learning business case, the data science and IT teams need to extensively collaboration in order to increase the pace of scaling and pushing of multiple machine learning models to production through continuous training, continuous validation, continuous deployment and continuous integration with governance. Machine Learning Operations (MLOps) has carved a new era of DevOps paradigm in the machine learning/artificial intelligence realm by automating end-to-end workflows.

As we are optimizing the models and bringing the data processing and analysis closer to the edge, data scientists and ML engineers are continuously finding new ways to push the complications involved with operationalization of models to such IoT (Internet of things) edge devices. 

A LinkedIn publication revealed that by 2025, the global AI spend would have reached $232 billion and $5 trillion by 2050. According to Cognilytica, the global MLOps market will be worth $4 billion by 2025. The industry was worth $350 million in 2019. [1]

Models running in IoT edge devices need to be very frequently trained due to variable environmental parameters, wherein continuous data drift and limited access to such IoT edge solutions may lead to degradation of the model performance over time. The target platforms on which ML models need to be deployed can also vary, such as IoT Edge or to specialized hardware such as FPGAs which leads to high level of complexity and customization with regards to MLOps on such platforms. 

Models can be packaged into docker image for the purpose of deployment post profiling the models by determining the cores, CPU and memory settings on said target IoT platforms. Such IoT devices also have multiple dependencies for packaging and deploying models that can be executed seamlessly on the platform. Hence, model packaging is easily implemented through containers as they can span over both cloud and IoT edge platforms.

When we are running on IoT edge platforms with certain dependencies of the device, a decision needs to be taken which containerized machine learning models need to be made available offline due to limited connectivity. An access script to access the model, invoke the endpoint and score the request incoming to the edge device needs to be operational in order to provide the respective probabilistic output. 

Continuous monitoring and retraining of models deployed in the IoT devices need to be handled properly using model artifact repository and model versioning features as part of MLOps framework. Different images of the models deployed will be stored in the shared device repository in order to quickly fetch the right image at the right time to be deployed to the Iot device. 

Model retraining can be triggered based on a job scheduler running in the edge device or when new data is incoming, invoking the rest endpoint of the machine learning model. Continuous model retraining, versioning and model evaluation become an integral part of the pipeline. 

In case the data is frequently changing which can be the case with such IoT edge devices or platforms, the frequency of model versioning and refreshing the model due to variable data drift will enable the MLOps engineer persona to automate the model retraining process, thereby saving time for the data scientists to deal with other aspects of feature engineering and model development. 

In time, the rise of such devices continuously collaborating over the internet and integrating with MLOps capabilities is poised to grow over the years. Multi-factored optimizations will continue to occur in order to make the lives of data scientists and ML engineers more focused and easier from model operationalization standpoint via an end-to-end automated approach.