The Linux Foundation has posted session descriptions for the Embedded Linux Conference Europe and OpenIoT Summit Europe, to be held Oct. 22-24, in Edinburgh, with topics ranging from RISC-V to deep learning to workplace diversity.
Even if you can’t make it to Edinburgh Oct. 22-24 for the Embedded Linux Conference Europe (ELCE) and co-located OpenIoT Summit Europe, the session descriptions are a good place to find clues about what’s hot in Linux and open source embedded technology. To be sure, the Linux Foundation offers a heavy dose of sessions on Linux Foundation projects such as Zephyr or Yocto Project, but it’s still a very inclusive collection from across the industry.
A different type of inclusivity is the unofficial theme of this year’s ELCE. There are a variety of sessions about best practices for increasing gender, ethnic, and other forms of diversity within technology firms and projects, including “Roadmapping an equitable open source movement,” “Building inclusive open source communities,” and “Opening doors: women in FOSS” (see farther below). A presentation by Mozilla Open Communities & Project Specialist Emma Irwin called “Metrics that matter for diversity & inclusion in open source” launches a three-day Diversity Empowerment Summit track.
A single registration covers both the ELCE/OpenIoT event and Open Source Summit Europe. It costs $800 through Aug. 18, after which it jumps to $950. The events will be directly followed by the Linux Security Summit (Oct. 25-26) and Zephyr Hackathon (Oct. 25).
Following the opening remarks from Linux Foundation Executive Director Jim Zemlin on Oct. 22, there will be an “Energy Panel” discussion and a yet to be announced Intel keynote speaker. This will be followed by the usual keynote discussion panel between Linux creator Linus Torvalds and long-time embedded Linux leader Dirk Hohndel, who is now a VP and Chief Open Source Officer at VMware.
There are also a few gaps in the Oct. 23 and 24 keynote schedules including an undisclosed Microsoft speaker on Oct. 23. The day’s keynotes will start with Vibrant Data’s Eric Berlow speaking on using Big Data software to analyze complex data to better understanding problems such as today’s Middle East conflicts. There’s also and keynotes from Johanna Koester of IBM and Patrick Ball of the Human Rights Data Analysis Group that appear to be related.
The Oct. 24 keynotes include Dr. Alexander Nitz of the Max Planck Institute on using the PyCBC Python library for studying gravitational waves. Other keynotes include LWN.net’s Jonathan Corbet, Mifos Initiative CEO Ed Cable, and game designer Brenda Romero.
The following is a somewhat arbitrary selection of some of the more intriguing presentations that follow, mostly from the ELCE track:
- Oct. 22:
- Supporting hardware codecs in a Linux system — Maxime Ripard, Bootlin
— Ripard details Bootlin’s project to develop a V4L2 driver for Allwinner processors on top of the mainline kernel.
- Opening Doors: women in FOSS — Jigyasa Grover, Univ. of Calif., San Diego — According to a recent survey by Github, only 3 percent of open source project participants are women. Grover offers a vision for how to change that.
- RISC-V ISA and Foundation overview — Rick O’Connor, RISC-V Foundation — Everything you need to know about the fast-growing open source RISC-V ISA. This will be followed by a 4:15 session from Comcast’s Khem Raj on using RISC-V with embedded Linux.
- ACRN: a big little hypervisor for IoT development — Eddie Dong, Intel — The lightweight, safety critical ACRN hypervisor could emerge as a major virtualization technology in embedded and automotive Linux projects such as AGL.
- Zephyr and Trusted Execution Environments — Andy Gross, Linaro — It’s good to see the Arm-financed Linaro discussing ARMv8M and ARMv7M TEE support in the open source Zephyr RTOS, which competes with Arm’s own Mbed.
- Embedded computer vision and machine learning for full-cycle drone autonomy — Mohammed Kabir, MIT — Kabir describes an MIT project for a sub-1kg “racer class” drone that combines an Intel Atom SoC, a Myriad 2 neural accelerator, and multiple cameras.
- Deep learning in OpenCV — Peng Li and Wu Zhiwen, Intel — Li and Zhiwen explain OpenCV’s DNN deep learning module, including pre-training with major ML frameworks like Caffe, TensorFlow, and torch, and network acceleration tips.
- The power-supply subsystem — Sebastian Reichel, Collabora — Untangling how the Linux kernel subsystems handles batteries and battery chargers.
- Comparison of voice assistant SDKs for embedded Linux devices — Leon Anavi, Konsulko Group — Pros and cons of Alexa, Google Assistant, and the open source Mycroft, with tips for prototyping on SBCs like the Raspberry Pi.
- Why embedded cameras are difficult, and how to make them easy — Laurent Pinchart, Ideas on Board — Pinchart reveals a new V4L2 sub-project to “develop a camera stack that will be open-source-friendly while still protecting vendor core IP.”
Full session descriptions and links to reservations for the Embedded Linux Conference Europe (ELCE) and OpenIoT Summit Europe, to be held Oct. 22-24, in Edinburgh, Scotland, may be found here.
>> Source Link