IoT on Spark

InfoObjects experience

Big Data Consulting with a focus on Apache Spark.

Industrial IoT Data

 

What is currently measured? What goes into the cost of unscheduled downtime:

%

Trying to figure out if there is a real problem

+

%

Diagnosing the problem

+

%

Finding resources to fix the problem

=

%

Time spent even before repairs have begun

In a paper mill, one hour of down time can cost over

1 Million Dollars

C

789+
456
123÷
0.=x

Run the numbers

So how much money are you wasting on a 4-hour down time?
Do the math. Here’s a calculator.

How do we fix it?

We Predict

when a part is going to fail by streaming real time sensor data and applying machine learning models

We Provide

real-time alarm and trouble shooting integration

We Engineer

end-to-end remote monitoring on high value assets

Data Ingestion/Architecture

 

  • Millions of Sensors
  • Millions of Events
  • Valocity
  • Volume
  • Store (audit, diagnosis)
  • Near Real-time Processing

Technology Choices

Data Store

  • IoT needs heavy writes and light reads
  • We design to customer requirements
  • AWS, Azure or on Prem

API Gateway

  • Hosted PaaS
  • Kafka Rest

Streaming Pipe

  • Kafka
  • Kinesis/EventHub

Contact us to learn more

Top