IoT and Cloud developers are in high demand today. Various platforms for IoT like Azure, AWS, GE Predix, Google and ThingsWorx are improving their service architectures and offerings to gain massive share in the market. In such situation it is always good to be certified and have your skills sorted out with multiple platforms. GE Predix is one of the Cloud which is specially designed to handle Industrial IoT applications right from managing railway operations to national level power generation grids. I must say that there are pretty good features in Predix which can be useful for industrial scenarios, but the operational layer is quite different than the traditional Public Clouds.
What is the exam all about?
GE Predix as a platform have two certifications at this point of time. One for the Solutions Architect and the other for Developers. The real action lies in the developer’s exam because it needs good amount of Predix platform knowledge with some hands-on for implementing couple of services and end-to-end machine to Cloud connectivity.
One needs to score around 71% to clear the exam. There are a total of 60 questions and 90 minutes are given to attempt all these questions. It is a remote proctored exam hence you can give this exam from home or office. The remote proctor will monitor the environment through camera and microphone. The exam is objective, and one must select the best option.
Predix as a Cloud for Developers
Predix developed by GE is a fascinating IIoT platform. All the services on this platform closely focus on industrial scenarios. There are many features like asset modeling, running analytics as modules on edge, running Postgre SQL as service, managing time series data and advanced UI mechanism for data visualization backed by Polymer and Angular JS making Predix a suitable platform for industrial internet applications. Predix Cloud is built on top of Cloud Foundry. So, developers who are aware of Cloud Foundry interfaces and architecture would feel at home!
The GUI for Predix doesn’t look mature and overwhelming like one for Azure or AWS. Most of the operations are done through CLI or code. There are few services which have free tier but most of the important data services requires a paid tier subscription and hence you need to provide your credit card details and spend a few dollars to perform your POCs.
Also, using Predix in real life could be bit tricky as they don’t have a very strong community support. Java is the key factor for using Predix. All the vital cogs are built using Java in Predix like Predix Machine Application, Analytic Modules, Visualization, etc. The documentation is extensive but still following the same and completing the hands-on will not be a cake walk. Every service will have ‘Getting Started’ section which is the actual hands-on guide for that service.
Predix prefers to use Eclipse in all development activities. Application hosting is just a push away from command-line. All the configuration files are mostly YAML.
Pre-requisites and ‘Good to have’ skills for the exam preparation
Understanding of JAVA is important
Fundamental working of APIs including header types, call functions, JSON responses
Basic knowledge of Docker would be useful as Predix Machine applications would need it
To operate Predix Cloud, one should know the fundamental architecture of Cloud Foundry. If this is ignored than the picture of ‘what’s going on’ would always be opaque. Understand the Predix Cloud components and start playing with cf interface. List and learn all the commands for logging in to your Predix Org, creating service instances, pushing applications for hosting, binding applications and deleting applications.
2. OAuth and Grant Types
Almost 20% of the questions are from security section which deals with UAA. UAA is the OAuth service to which every application binding is done for securing the access. Spending time in understanding the working of OAuth and different grant types with the perspective of Predix UAA is very vital. The structure of tokens, the process of fetching the tokens and protocols involved in various scenarios builds a good security knowhow of the platform.
3. Integration of Services and Flow of Data
An IoT solution would include machines, edge, ingestion, stream analytics, data storage and visualization as standard components. Learn the use cases of all relevant services and find out how services are integrated, which services could possibly communicate directly with each other and how is it made possible. One should be capable enough to send, process, store and visualize sensor data generated from an on-premises resource like Raspberry Pi of BeagleBone. In real life, such devices get replaced by other gateways, PLCs, SCADA systems, etc.
4. API Reference
A detailed API reference documentation is provided for every service. It is very common to use APIs for almost every kind of action which we want to take on Cloud based resource. For performing all common tasks like creation, modification, deleting, binding, exposing a resource, there would be different API calls to make. Understanding their structure, headers, response values and security methods is core for every service.
5. Asset Modeling
Predix is built for managing complex machinery data of factories, automobiles, energy grids, healthcare management systems and aerospace systems. These systems have complicated inter-relationship and hierarchical set of components. Asset Modeling service helps to manage this complication. Every asset is identified uniquely. For monitoring these assets and defining schema of machineries we can use Asset service.
There are too many things to learn. If you don’t have a background with Industrial IoT and using Cloud for cyber physical systems, we can help you with our well-designed training program i.e. GE Predix Developer Certification BootCamp to save your precious time and prepare for this exam.
Best of luck and please share your valuable comments or any questions. We will get back on that as soon as possible.