Internet of Things has shown a remarkable improvement over the years. The basic IoT embedded devices with sensors, have now advanced to a level where AI can be deployed into IoT devices to make them smarter. The IoT World 2018 conference was held from May 14th to 17th, at Santa Clara Convention Center, CA, USA. Al special developer centric conference designed specifically for technologists was also part of the larger conference.
The agenda for the developers’ conference was to bring together the technical leaders who have contributed with their innovations in the IoT market and tech enthusiasts who look forward to develop their careers in this domain.This conference also included sessions such as SAP learning, and interesting keynotes on Intelligent Internet of things.
Here are five sessions that caught our eyes at the developer conference in IoT World 2018.
How to develop Embedded Systems by using the modern software practices, Kimberly Clavin
Kimberly Clavin highlighted that a major challenge in developing autonomous vehicles include, system integration and validation techniques. These techniques are used to ensure quality factor within the code.
There are a plethora of companies that have software as their core and use modern software practices such as (Test Driven Development)TDD and Continuous Integration(CI) for successful development. However, the same tactics cannot be directly implemented within the embedded environment.
Kimberly presented ways to adapt these modern software practices for use within the development of embedded systems. This can help developers to create systems that are fast, scalable, and a cheaper. The highlights of this session include,
- Learning to test drive an embedded component.
- Understanding how to mock out/simulate an unavailable component.
- Application of Test Driven Development (TDD), Continuous Integration (CI) and mocking for achieving a scalable software process on an embedded project.
How to use Machine Learning to Drive Intelligence at the Edge, Dave Shuman and Vito De Gaetano
Edge IoT is gaining a lot of traction of late. One way to make edge intelligent is by building the ML models on cloud and pushing the learning and the models onto the edge. This presentation session by Dave Shuman and Vito De Gaetano, shows how organizations can push intelligence to the edge via an end-to-end open source architecture for IoT.
This end-to-end open source architecture for IoT is purely based on Eclipse Kura and Eclipse Kapua. Eclipse Kura is an open source stack for gateways and the edge, whereas Eclipse Kapua is an open source IoT cloud platform. The architecture can enable:
- Securely connect and manage millions of distributed IoT devices and gateways
- Machine learning and analytics capabilities with intelligence and analytics at the edge
- A centralized data management and analytics platform with the ability to build or refine machine learning models and push these out to the edge
- Application development, deployment and integration services
The presentation also showcased an Industry 4.0 demo, which highlighted how to ingest, process, analyze data coming from factory floors, i.e from the equipments and how to enable machine learning on the edge using this data.
How to build Complex Things in a simplified manner, Ming Zhang
Ming Zhang put forth a simple question,“Why is making hardware so hard?” Some reasons could be:
- The total time and cost to launch a differentiated product is prohibitively high because of expensive and iterative design, manufacturing and testing.
- System form factors aren’t flexible — connected things require richer features and/or smaller sizes.
- There’s unnecessary complexity in the manufacturing and component supply chain.
Designing a hardware is a time-consuming process, which is cumbersome and not a fun task for designers, unlike software development. Ming Zhang showcased a solution, which is ‘The zGlue ZiPlet Store’ — a unique platform wherein users can build complex things with an ease. The zGlue Integrated Platform (ZiP) simplifies the process of designing and manufacturing devices for IoT systems and provides a seamless integration of both hardware and software on a modular platform.
Building IoT Cloud Applications at Scale with Microservices, Dave Chen
This presentation by Dave Chen includes how DConnectivity, big data, and analytics are transforming several business types. A major challenge in the IIoT sector is the accumulation of humongous data. This data is generated by machineries and industrial equipments such as wind turbines, sensors, and so on. Valuable information out of this data has to be extracted securely, efficiently and quickly.
The presentation focused on how one can leverage microservice design principles and other open source platforms for building an effective IoT device management solution in a microservice oriented architecture. By doing this, managing the large population of IoT devices securely becomes easy and scalable.
Design Patterns/Architecture Maps for IoT Design patterns are the building blocks of architecture and enable developers and architects to reuse solutions to common problems.
The presentation showcased how various common design patterns for connected things, common use cases and infrastructure, can accelerate the development of connected device.
Extending Security to Low Complexity IoT Endpoint Devices, Francois Le
At present, there are millions of low compute, low power IoT sensors and devices deployed. These devices and sensors are predicted to multiply to billions within a decade. However, these devices do not have any kind of security even though they hold such crucial, real-time information. These low complexity devices have:
- Very limited onboard processing power,
- Less memory and battery capacity, and
- are typically very low cost.
Low complexity IoT devices cannot work similar to IoT edge device, which can easily handle validation and encryption techniques, and also have huge processing power to handle multiple message exchanges used for authentication.
The presentation states that a new security scheme needs to be designed from the ground up. It must acquire lesser space on the processor, and also have a low impact on battery life and cost. The solution should be:
- IoT platform agnostic and
- Easy to implement by IoT vendors,
- Easily operated over any wireless technologies (e,g, Zigbee, BLE, LoRA, etc.) seamlessly
- Transparent to the existing network implementation.
- Automated and scalable for very high volumes,
- Evolve with new security and encryption techniques being released
- Last for a long time in the field with no necessity to update the edge devices with security patches.
Apart from these, many other presentations were showcased at the IoT World 2018 for developers. Some of them include, Minimize Cybersecurity Risks in the Development of IoT Solutions, Internet of Things (IoT) Edge Analytics: Do’s and Don’ts.
Read more keynotes presented at this exciting IoT World conference 2018 on their official website.
- Cognitive IoT: How Artificial Intelligence is remoulding Industrial and Consumer IoT
- How IoT is going to change tech teams
- AWS Greengrass brings machine learning to the edge