| © Copyright 2024, InfluxData
1
Real Time Monitoring
& Predictive
Maintenance in
Aerospace Tech
Suyash Joshi - sjoshi@influxdata.com
Sr. Software Developer Advocate
InfluxData
| © Copyright 2024, InfluxData
2
Aerospace tech “lives & breathes” time series data
• Measuring change over time – status, performance, problems
• Events and activity over various time intervals
• Time & Data is often the magnifying glass for MRO related
Telemetry for Analytics and Predictive Maintenance
| © Copyright 2024, InfluxData
3
Monitoring & Analytics of the entire lifecycle
Manufacturing
100,000+
Sensors
Development &
Testing
Billions
Data Points
In-flight
Operations
Billions
Device Tags
<1 second
Data Frequency
Take off &
Landing
| © Copyright 2024, InfluxData
4
Challenges with Monitoring Huge Data
Data is continuously
arriving at high speed
and volume
Applications must
analyze data within
streams and act in real
time
Higher number of tags
collected cause high
cardinality impacting
performance
Massive Scale Real Time Action Data Cardinality
| © Copyright 2024, InfluxData
5
Time Series Database
● Scale
Designed to scale for large volumes of
time series data
● Distributed
Non-blocking high volume writes and
reads
● Availability
Write and read availability are prioritized
over consistency
● Management
Data lifecycle management with built-in
data retention
● Flexible
Schema on write
Unlimited
Data points per second
(high ingestion)
Billions
Data cardinality
(single datastore for all time
series dat)
Hot data in memory
Real Time
Lowest cost
storage
Cold data in object store
| © Copyright 2024, InfluxData
6
TIME SERIES
Evolution of Databases
DOCUMENT SEARCH
RELATIONAL
• Events, metrics, time-stamped
• For IoT, analytics, cloud native
• Distributed
search
• Logs
• Geo
• High
throughput
• Large
document
• Orders
• Customers
• Records
Time series is fastest growing
data category by far
All others
Time series
source: DB Engines
influxdb
| © Copyright 2024, InfluxData
7
InfluxDB 3: Run on cloud & on-premises
Core (Alpha Release)
- Open Source (Apache 2.0/MIT)
Enterprise
- Alpha Release (GA in April, 2025*)
Cloud Serverless
• managed service for small & medium
workloads
Cloud Dedicated
• managed service for large enterprise
workloads
Clustered
• software for large enterprise workloads in
self-managed environments
| © Copyright 2024, InfluxData
8
Use case architecture | Aerospace Monitoring
| © Copyright 2024, InfluxData
9
Edge Data Replication (EDR)
EDGE
EDR removes a lot of complexity to
maintain Edge to Cloud replication and
thus enabling Customers to bring OT
and IT closer and eliminate data
silos.
HUB
| © Copyright 2024, InfluxData
10
Aviation MRO (Maintenance, Repair, Overhaul)
| © Copyright 2024, InfluxData
11
Forecasting using Machine Learning
Statistical & ML Methods
1. Naive Method
● Linear Regression: Predicts future values using a straight-line trend.
● Random Walk: Assumes future values change randomly, like stock prices.
2. Exponential Smoothing
● Gives more weight to recent data and less to older data for better predictions.
3. Lagging Features
● Uses past values to predict the future, capturing time-based patterns (e.g., sales or
temperature).
4. ARIMA
● AR (Autoregressive): Predicts future values based on past values.
● MA (Moving Average): Smooths past errors to improve predictions.
5. LLMs
● Time Series LLMs that come pre-trained on large data sets
| © Copyright 2024, InfluxData
12
12
Useful TimeSeries ML Libraries
Unsupervised and rule-based
time series anomaly detection
The ADTK package allows you
to easily build an effective
detection model from a variety
or rule-based anomaly
detection methods.
“Prophet is a procedure for
forecasting time series data
based on an additive model
where non-linear trends are fit
with yearly, weekly, and daily
seasonality, plus holiday
effects.”
A Neural Network based
Time-Series model, inspired by
Facebook Prophet and AR-Net
(Autoregressive neural
network), built on PyTorch.
“A new powerful open source
library to perform time series
forecasting etc using LLM by
Amazon Research”
Anomaly Detection Toolkit
(ADTK)
FB Prophet Neural Prophet Chronos
| © Copyright 2024, InfluxData
13
AST Space Mobile migrated from PostgreSQL
to InfluxDB 3.0 to power their long term
telemetry store for satellite constellation
STORY:
Mission is to connect every person on the planet
with a high-speed internet from almost anywhere
in the world
PROBLEM:
Collect, analyze & store 100K+ metrics per
second from hundreds of satellites
INFLUXDB VALUE:
• Fast query responses for queries over long
time ranges
• Cost-effective support for data retention over
10+ years
Satellite Based
Communications
| © Copyright 2024, InfluxData
14
Problem: data historian and machine-specific
monitoring tools can’t handle data volume
- Potential for millions of dollars in aircraft
repair work, scrapped work, and delayed
delivery due to tiny variations in conditions
- AVEVA Wonderware + machine-specific
monitoring failed to keep up volume and
query performance.
- Unable to provide the factory-wise view
InfluxDB results:
- Team is now monitoring more than 3000
parameters per second in one factory-wide
view
- Met stringent security and compliance
requirements for both commercial and
military clients
Aircraft Manufacturing
| © Copyright 2024, InfluxData
15
FlightAware Demo!
| © Copyright 2024, InfluxData
16
Thank You!
Resources
● Predictive Maintenance in Aerospace Magazine Article (Page 58)
● Learn to Forecast TimeSeries Data: influxdata.com/blog/forecast-time-series-data-ml-influxdb
● Flight Tracker Demo: https://github.com/InfluxCommunity/Flight-Demo
● Blogs: https://www.influxdata.com/blog
● Documentation: https://docs.influxdata.com
● InfluxDB University (free training): https://influxdbu.com
● Community: https://influxcommunity.slack.com & https://community.influxdata.com
❖ www.influxdata.com/cloud (Free to try!)
❖ via cloud marketplace
| © Copyright 2024, InfluxData
17
T H A N K Y O U

InfluxDB Presentation for Aerospace 2025 Conference

  • 1.
    | © Copyright2024, InfluxData 1 Real Time Monitoring & Predictive Maintenance in Aerospace Tech Suyash Joshi - sjoshi@influxdata.com Sr. Software Developer Advocate InfluxData
  • 2.
    | © Copyright2024, InfluxData 2 Aerospace tech “lives & breathes” time series data • Measuring change over time – status, performance, problems • Events and activity over various time intervals • Time & Data is often the magnifying glass for MRO related Telemetry for Analytics and Predictive Maintenance
  • 3.
    | © Copyright2024, InfluxData 3 Monitoring & Analytics of the entire lifecycle Manufacturing 100,000+ Sensors Development & Testing Billions Data Points In-flight Operations Billions Device Tags <1 second Data Frequency Take off & Landing
  • 4.
    | © Copyright2024, InfluxData 4 Challenges with Monitoring Huge Data Data is continuously arriving at high speed and volume Applications must analyze data within streams and act in real time Higher number of tags collected cause high cardinality impacting performance Massive Scale Real Time Action Data Cardinality
  • 5.
    | © Copyright2024, InfluxData 5 Time Series Database ● Scale Designed to scale for large volumes of time series data ● Distributed Non-blocking high volume writes and reads ● Availability Write and read availability are prioritized over consistency ● Management Data lifecycle management with built-in data retention ● Flexible Schema on write Unlimited Data points per second (high ingestion) Billions Data cardinality (single datastore for all time series dat) Hot data in memory Real Time Lowest cost storage Cold data in object store
  • 6.
    | © Copyright2024, InfluxData 6 TIME SERIES Evolution of Databases DOCUMENT SEARCH RELATIONAL • Events, metrics, time-stamped • For IoT, analytics, cloud native • Distributed search • Logs • Geo • High throughput • Large document • Orders • Customers • Records Time series is fastest growing data category by far All others Time series source: DB Engines influxdb
  • 7.
    | © Copyright2024, InfluxData 7 InfluxDB 3: Run on cloud & on-premises Core (Alpha Release) - Open Source (Apache 2.0/MIT) Enterprise - Alpha Release (GA in April, 2025*) Cloud Serverless • managed service for small & medium workloads Cloud Dedicated • managed service for large enterprise workloads Clustered • software for large enterprise workloads in self-managed environments
  • 8.
    | © Copyright2024, InfluxData 8 Use case architecture | Aerospace Monitoring
  • 9.
    | © Copyright2024, InfluxData 9 Edge Data Replication (EDR) EDGE EDR removes a lot of complexity to maintain Edge to Cloud replication and thus enabling Customers to bring OT and IT closer and eliminate data silos. HUB
  • 10.
    | © Copyright2024, InfluxData 10 Aviation MRO (Maintenance, Repair, Overhaul)
  • 11.
    | © Copyright2024, InfluxData 11 Forecasting using Machine Learning Statistical & ML Methods 1. Naive Method ● Linear Regression: Predicts future values using a straight-line trend. ● Random Walk: Assumes future values change randomly, like stock prices. 2. Exponential Smoothing ● Gives more weight to recent data and less to older data for better predictions. 3. Lagging Features ● Uses past values to predict the future, capturing time-based patterns (e.g., sales or temperature). 4. ARIMA ● AR (Autoregressive): Predicts future values based on past values. ● MA (Moving Average): Smooths past errors to improve predictions. 5. LLMs ● Time Series LLMs that come pre-trained on large data sets
  • 12.
    | © Copyright2024, InfluxData 12 12 Useful TimeSeries ML Libraries Unsupervised and rule-based time series anomaly detection The ADTK package allows you to easily build an effective detection model from a variety or rule-based anomaly detection methods. “Prophet is a procedure for forecasting time series data based on an additive model where non-linear trends are fit with yearly, weekly, and daily seasonality, plus holiday effects.” A Neural Network based Time-Series model, inspired by Facebook Prophet and AR-Net (Autoregressive neural network), built on PyTorch. “A new powerful open source library to perform time series forecasting etc using LLM by Amazon Research” Anomaly Detection Toolkit (ADTK) FB Prophet Neural Prophet Chronos
  • 13.
    | © Copyright2024, InfluxData 13 AST Space Mobile migrated from PostgreSQL to InfluxDB 3.0 to power their long term telemetry store for satellite constellation STORY: Mission is to connect every person on the planet with a high-speed internet from almost anywhere in the world PROBLEM: Collect, analyze & store 100K+ metrics per second from hundreds of satellites INFLUXDB VALUE: • Fast query responses for queries over long time ranges • Cost-effective support for data retention over 10+ years Satellite Based Communications
  • 14.
    | © Copyright2024, InfluxData 14 Problem: data historian and machine-specific monitoring tools can’t handle data volume - Potential for millions of dollars in aircraft repair work, scrapped work, and delayed delivery due to tiny variations in conditions - AVEVA Wonderware + machine-specific monitoring failed to keep up volume and query performance. - Unable to provide the factory-wise view InfluxDB results: - Team is now monitoring more than 3000 parameters per second in one factory-wide view - Met stringent security and compliance requirements for both commercial and military clients Aircraft Manufacturing
  • 15.
    | © Copyright2024, InfluxData 15 FlightAware Demo!
  • 16.
    | © Copyright2024, InfluxData 16 Thank You! Resources ● Predictive Maintenance in Aerospace Magazine Article (Page 58) ● Learn to Forecast TimeSeries Data: influxdata.com/blog/forecast-time-series-data-ml-influxdb ● Flight Tracker Demo: https://github.com/InfluxCommunity/Flight-Demo ● Blogs: https://www.influxdata.com/blog ● Documentation: https://docs.influxdata.com ● InfluxDB University (free training): https://influxdbu.com ● Community: https://influxcommunity.slack.com & https://community.influxdata.com ❖ www.influxdata.com/cloud (Free to try!) ❖ via cloud marketplace
  • 17.
    | © Copyright2024, InfluxData 17 T H A N K Y O U