Sample Scramjet architecture for Internet of Things (IoT)
The diagram below shows collecting and processing metrics and events from IoT devices. Scramjet Cloud Platform performs events processing for triggering alarms and detecting anomalies by AI/ML algorithms.
Use Case: Collect data from weather stations
Scramjet Cloud Platform can gather data from multiple weather and environmental sensors - even a single sequence with input endpoint working on our engine can handle data from multiple sensors with very minimal resources consumption and high performance. The sequence is also able to transform data on the fly, extract only required information, and persist it to various databases or other data stores. No need to build complex infrastructure, you can handle multiple sensors and save only the data you need, saving implementation infrastructure and data storage costs.
Example:
The City governor's office is deploying a set of smog sensors across various locations in the city. This project is supported by the city budget and appraised by citizens' votes; however, hardware purchase will consume most of the project's budget. The office is looking for a simple, low-cost, self-maintaining efficient solution that will help to collect data and save it to a local server for further analysis and creation of dashboards.
Scramjet Cloud Platform can deliver here both data ingestion, data extraction, and saving data to storage without the complexity of building a custom solution from multiple software modules.
Use Case: React to device malfunction or security alarms
Production and logistics companies now use a complex set of sensors, alarms, detectors to manage and reduce risks related to their production and logistics locations.
Scramjet Cloud Platform can be used as both an entry point, event handler, and an alarm generator for a set of alarms, such as security alarms, device malfunction alarms, or environmental alarms.
Example:
An ice cream production company decided to internally deploy a set of alarms in their large warehouse, distribution, and logistics building. These sensors perform various roles: guard entry to the facility, perform smoke detection, monitor devices (such as cooling devices), check parameters such as temperature and humidity. This set of sensors generates a large volume of data. The company is looking for a simple to manage and economical solution to collect, pre-process these data and send alarm notifications (such as SMS or email) to proper teams if needed.
Scramjet Cloud Platform can be used as a single solution to cover all these requirements - a set of sequences will be able to consume data streams from different devices, check values inside messages, and trigger alarm notifications in proper channels. All this can be achieved with one platform and several, simple to write applications (sequences).
Use Case: Collect live data from monitoring drones
High-tech companies delivering advanced drone technology, such as industry environmental monitoring drones or search and rescue drones might need to perform a live analysis of captured images and the drone control center should react to patterns on these images in real-time, issuing warnings and alarms.
Scramjet Cloud Platform instances process video streams the same way as text streams and either: pre-process images for ML algorithms that will perform inference (for example, detecting lost child in the forest) or use ML trained model, embedded directly into the instance and performing analysis of video frames on the fly.
Example:
A drone systems vendor would like to extend the capabilities of their monitoring drones and perform near-real-time analysis of the monitored area and instant alarm invocation. The previous POC failed to meet requirements: sending a live stream of video to a local control station via GSM connection in difficult terrain was much too slow and error-prone. The team decided to look for a solution that will allow them to pre-process video on the drone and send only selected video frames for further analysis.
Scramjet can provide both elements of this solution: Scramjet Transform Hub can be installed on the drone itself and can pre-process video stream and send only every n-th of a video frame to Scramjet Cloud Platform for ML- images patterns analysis and triggering alarms, if needed. The whole management of data streams, both on the drone and in the cloud, is done by Scramjet Cloud Platform. SCP works here as a data orchestrator, deployment, and runtime environment.