To make it more interesting I thought it’d be cool to restrict the machine learning model to a single type of item and add support for voice control to change between types. While I was at it I added the audio from the Portal turrets, because why not?
Last night the object detection project, which is using a Raspberry Pi model 3b+ and an Intel / Movidius Neural Compute Stick 2 (NCS2), finally started working as expected. The video feed comes from a Pi cam v2.1. Kudos for Adrian at Pyimagesearch for his very helpful guide on this topic!
After a fresh install of Ubuntu server 20.04 on a Raspberry Pi (3b+ in this case) you may find that the default login (user: “ubuntu” password: “ubuntu”) won’t get you logged in.
After waiting for a few minutes while googling this the Rpi output some cloud-init and SSH-key messages. When trying again the default ubuntu/ubuntu login worked just fine.
So, it turns out that even though the login screen is shown and it looks like the OS has fully booted, it won’t be possible to log in until the cloud-init etc. have finished (which may take 5min or so)!
After having received several inquiries from people about how to get started with EdgeX Foundry I’ve decided to write a hands-on tutorial. Hopefully this will make it easier for newcomers to setup a system, configure data ingestion, data export & many other things.
Incidentally, getting started with EdgeX Foundry is also an excellent way to learn how to practically leverage new concepts and technologies in IT. In addition to learning about open source IoT solutions, the guide also covers topics like:
- Building, running and monitoring containers
- Grouping containers with docker-compose
- REST APIs
- Python scripting
- Tools like Postman, cURL, etc.
The guide started as a blog post but ended up being way too long. Now it’s in PDF format and clocks in at 48 pages. Hopefully not too long for those looking to get started 🙂
In early 2020 a new feature was added to the PowerEdge 14G servers called “Telemetry Streaming’. This feature makes it possible to send a continuous stream of telemetry containing in-depth information about the state of the server and its various components including, but not limited to, the following:
- CPU, Memory and Fans
- FPGA and GPU
- PCIe slots
- Airflow inside server
- Power usage information
Since the level and depth of information collected with this method FAR exceeds what has been previously possible using IPMI or other tools, this feature can help in several areas. For example:
- Power ML algorithms for Anomaly Detection
- Provide detailed inventory, usage and status information
- Assist with security and auditing
Blog posts in this series
- Introduction and overview (this post)
- Enabling telemetry streaming (3 methods of enabling Tel. Streaming)
- Configuring telemetry streaming (setup of InfluxDB, Grafana, etc.)
Introductory video and demo
Links to useful documents / scripts published elsewhere:
Implementation of the DMTF Redfish API on Dell EMC PowerEdge Servers: http://en.community.dell.com/techcenter/extras/m/white_papers/20442330
Python scripts on Github: https://github.com/dell/iDRAC-Telemetry-Scripting
Introduction to Telemetry Streaming (highly recommended practical guide!): https://downloads.dell.com/Manuals/Common/dell-emc-idrac9-telemetry-streaming-basics.pdf
Analyst report on the Telemetry Streaming feature: https://www.dellemc.com/resources/en-us/asset/analyst-reports/products/servers/tolly-dell-emc-idrac9-v4-telemetry-streaming.pdf