Overview
On 14th June 2023 we had our first IoT Hackathon at the International User Group Conference 2023 in Budapest.
Btw. if you interested in the recordings of the IUG 2023 session check this link out.
In total 23 participants of companies Kunbus, IBM,Simplias, Novatec, WIKA, A1 Digital and Software AG worked in 6 mixed teams on 6 specific topics
Ideation & Team Setup
After a short introduction the teams started the ideation and picked one of 13 challenges or developed new ideas they wanted to work on.
They came up with 6 teams and 6 topics they wanted to work on that day. One of them has been awarded as the “winner” of the Hackathon and brought home I unique trophy + RevPis sponsered by Kunbus!
Let me share the key results of the teams with you!
Team Notification 3.0
Team Notification 3.0 picked the blueprint to manage subscriptions used in the Notification 2.0.
Currently there is no UI available to manage all subscription in tenants. As subscriptions may consume a lot of resources, especially if no client is connected to consume the messages, managing subscriptions help to reduce costs. In multi tenant solutions it is even more necessary to see which customer tenants have how many subscriptions and consume a lot of resources.
In the end they delivered a new UI plugin that can be added to the Administration UI to manage all the subscription in one tenant
From that UI you can delete existing subscriptions opening a confirmation dialog
Team SwAGson
Team SwAGson mainly consists of IBM & Software AG participants. Goal was to train IBM Watson with the Cumulocity IoT documentation and integrate IBM Watsons Assistant as a chatbot similar UI extension to Cumulocity IoT. Stretch goals were to integrate further sources like Tech Community.
The result were quite impressive. IBM Watson Assistant was trained for some default questions like “How to connect a sensor?”. Other more complicated questions are directly sent to the IBM Watson Neural Seek.
The Chatbot is directly integrated to Cumulocity IoT using a plugin and enhancing existing application like the cockpit in this case.
Team Tapedeck
Team Tapedeck came up with their idea directly on the Hackathon. They addressed the need of real productive field data for simulation and testing of a device fleet. Very often simulated devices using synthetic data. Wouldn’t it be nice to just record “real” data of a device and “replay” that data on multiple other devices? Team tapedeck achieved exactly that using thin-edge.io!
Firstly you can trigger the recording for a period of time. thin-edge.io will record all data for that period and store it. That snapshot of data can be “played” on multiple other devices spawning containers with the before collected data. So you can simulate a whole fleet with realistic data.
Team Tapedeck pitching their solution
Team BigBrother
Team BigBrother actually combined two blueprints: Track anything & Smart Device Fleet Management.
First of all they used go-c8y-cli to simulate a lot of devices and changing their position in a desired area. They integrated a barcode scanner to simulate the scanning of any asset. By scanning the Id of the asset the position of the scanner will be added to the asset information in Cumulocity IoT and visualized on a map.
To model the asset the digital twin manger was used. Position data are correlated to the assets using Streaming Analytics and unique IDs.
For that complexity and usage of multiple tools the expert jury voted this solution as the Winner! of the Hackathon!
Team Alex
Team Alex mainly was @apes who supported the other teams and used some “free” time to develop an opcua extensions for go-c8y-cli.
With that extension you can use the go-c8y-cli to manage your opcua servers using the command-line only. You can create & retrieve opcua server or even can start the address space scan.
This all has been achieved by just configuring yml files for the go-c8y-cli following the extension concept.
Team M&M
Team M&M were @Murat_Bayram and @Martin_Prodanov who used the hackathon to extend the Cumulocity IoT core with new aggregation methods for the measurements series API. Currently only min + max aggegration modes are supported by the series API. Team M&M enhanced the REST API to also return other aggregation values like “avg”, “sum” and “std”. For that they needed to extend the existing REST API but also change the UI.
What else happened?!
Beside that we all learned a lot that day, we also had a lot of fun!
Thanks to Kunbus who brought a connected beer barrel & connected coffee machine.
After the hackathon we enjoyed our time with some slice of a pizza and a glass of beer!
This article is part of the TECHniques newsletter blog - technical tips and tricks for the Software AG community. Subscribe to receive our quarterly updates or read the latest issue.