IoT Startup

The Challenge

Client provides solutions to manufacturing industries; especially textile ones in India, China and Vietnam. The basic premise is to provide a hybrid IoT solution which takes care of device as well as cloud side. Thinking Hut worked as a strategic partner, involved in ideation, development, QA and performance testing of the software part.

Challenges:

  1. A huge amount of data being generated continuously which raises a question of which data is meaningful to be sent and stored on the cloud. If the whole data is sent to the cloud it would consume large bandwidth. If meaningful data gets filtered out on the device side, notifications which call for action, would not get sent.
  2. Any data indicating an emergency and urgent attention at the machine on the factory flow should send a notification to the respective persons with least delay.

The Solution

  • We developed gateway apps for data collection and sending over the internet using the least bandwidth & costs. The cloud application provided batch analytics, real-time presentation of data, real-time analytics & aggregation of the data, event correlation, notification and current batch parameters as against the golden batch ones.
  • Developed & Tested gateway apps on Cisco IR829 and IR809 Industrial Integrated Services Routers.
  • Algorithms to suggest the parameters for ‘golden batch’ of production.
  • Factory floor – sensor data is collected optimally.
  • Data filtering & compression done for sending data over the internet using least bandwidth & costs.
  • MQTT over WebSocket used as a publish-subscribe model to send data from the edge to cloud.
  • Visualization and reporting of the analyzed data.

The Outcome

  • High alerts and notifications are being sent to the concerned personnel. This helps them to take action well in time and avoid the breakdowns. It reduces the downtime of the machines and of the production lines.
  • Higher productivity and efficiency of the production line.
  • Predictive and prescriptive analytics help to schedule machine maintenance.
  • Data filtering techniques helped in reducing the storage consumption and hence storage costing on the cloud.