Deconstructing the Data Test Track Derby

By Roguen Keller posted 10-02-2019 16:27


High-Level Architecture


The Data Test Track Derby architecture was designed with a few core principles in mind, such as:

  • Spin-up from anywhere, even in the absence or questionable internet connectivity
  • Show case convergence of IoT and smart data capture leveraging a handful of Hitachi software tools and platforms
  • Automated and enabled conclusions from the data capture
  • Open-source project demonstrating Jive integration best practices using Add-ons


The general flow for the integration runs as the following:

  1. Racers build their derby cars, register themselves and then the car before visiting the Derby start gate.
  2. Racers are assigned a Racer ID
    • This ID corresponds to their ID provided during racer registration
  3. Start gate assistant uses a simple UI to map users to respective lanes on the track, and starts the race.
  4. IoT sensors and other devices capture race speed information, and send racer profile information, lane assignments, race recording and speed details to HCP and local storage environments, or native cloud apps for further processing and orchestration.
  5. Pentaho tools processes the race result information and create reports and outcomes
  6. After the event, the environment can be decommissioned without any infrastructure mess

Local Environment

The Data Test Track Derby local environment represents all the sensors and hardware needed to collect the race details.


Raspberry Pi 3 (Adafruit)

  • Running Rasbian Jessie
  • 32GB microSD Card

  • We use NGROK as an on-the-ground option give us reliable connectivity regardless of network.  As long as the Raspberry Pi can see the internet, we can find the Raspberry Pi to manage access.


Related Resources:


Electronic Derby Timer w/Solenoid (Derby Magic)


Captures the run time on each lane


Related Resources:


Raspberry Pi Camera Board v2 - 8 Megapixels (Adafruit)



Used for registration, GraphicsMagick, ImageMagick (see README file)


Related Resources:


Pi Traffic Light (Amazon)


Provides a physical visual queue for various status during the Derby.  Also, goes well with the concept of racing, as we use it to simulate a race tree!


Related Resources:



IR Break Beam Sensor - 5mm LEDs x 2 (Adafruit)


Places alongside the track to capture race splits for various distances. (optional)


Related Resources:


Thunderboard Sense (Silicon Labs)


Connect to a Thunderboard Sense, poll device for measurement updates, and then proxy those to AWS IoT.


Related Resources:

See you at NEXT 2019!