Crowd Funding

264 Articles

AI In A Box Envisions AI As A Private, Offline, Hackable Module

October 30, 2023 by 10 Comments

[Useful Sensors] aims to embed a variety of complementary AI tools into a small, private, self-contained module with no internet connection with AI in a Box. It can do live voice recognition and captioning, live translation, and natural language conversational interaction with a local large language model (LLM). Intriguingly, it’s specifically designed with features to make it hack-friendly, such as the ability to act as a voice keyboard by sending live transcribed audio as keystrokes over USB.

Based on the RockChip 3588S SoC, the unit aims to have an integrated speaker, display, and microphone.

Right now it’s wrapping up a pre-order phase, and aims to ship units around the end of January 2024. The project is based around the RockChip 3588S SoC and is open source (GitHub repository), but since it’s still in development, there’s not a whole lot visible in the repository yet. However, a key part of getting good performance is [Useful Sensors]’s own transformers library for the RockChip NPU (neural processing unit).

The ability to perform things like high quality local voice recognition and run locally-hosted LLMs like LLaMa have gotten a massive boost thanks to recent advances in machine learning, and it looks like this project aims to tie them together in a self-contained package.

Perhaps private digital assistants can become more useful when users can have the freedom to modify and integrate them as they see fit. Digital assistants hosted by the big tech companies are often frustrating, and others have observed that this is ultimately because they primarily exist to serve their makers more than they help users.

Continue reading “AI In A Box Envisions AI As A Private, Offline, Hackable Module” →

The ‘Scope Of This Kickstarter? Ten Years.

September 4, 2023 by 18 Comments

It may have taken ten years to come through on this particular Kickstarter, but a promise is a promise. In late August 2023, backers who had since likely forgotten all about the project started receiving their oscilloscope watches from creator [Gabriel Anzziani]. Whatever the reason(s) for the delay, the watch looks great, and is miles ahead of the prototype pictures.

As you may have guessed, it functions as both a watch and an oscilloscope. The watch has 12- and 24-hour modes as well as an alarm and calendar, and the ‘scope has all the features of the Xprotolab dev board, which [Gabriel] also created: ‘scope, waveform generator, logic analyzer, protocol sniffer, and frequency counter.

Internally, it has an 8-bit Xmega microcontroller which features an internal PDI, and the display is a 1.28″ E ink display. When we covered this ten years ago, the screen was the type of Sharp LCD featured in the Pebble watch. [Gabriel]’s ‘scope watch features eight buttons around the edge which are user-programmable. One of [Gabriel]’s goals was for people to make their own apps.

Of course, the Kickstarter rewards are no longer available, but if you want to build your own small, digital ‘scope, check out this DIY STM32 project.

Image via the Company Formerly Known As Twitter

Pixel Pump, The Open Source Vacuum Pickup Tool Is Now Shipping

August 1, 2023 by 10 Comments

The Pixel Pump is an open source manual pick & place assist tool by [Robin Reiter], and after a long road to completion, it’s ready to ship. We first saw the Pixel Pump project as an entry to the 2021 Hackaday Prize and liked the clean design and the concept of a completely open architecture for a tool that is so valuable to desktop assembly. It’s not easy getting hardware off the ground, but it’s now over the finish line and nearly everything — from assembly to packaging — has been done in-house.

Pixel Pump with SMD-Magazines, also using foot pedal to control an interactive bill of materials (BoM) plugin.

Because having parts organized and available is every bit as important as the tool itself, a useful-looking companion item for the Pixel Pump is the SMD-Magazine. This is a container for parts that come on SMD tape rolls. These hold components at an optimal angle for use with the pickup tool, and can be fixed together on a rail to create project-specific part groups.

A tool being open source means giving folks a way to modify or add features for better workflows, and an example of this is [Robin]’s suggestion of using a foot pedal for hands-free control of the interactive BoM plugin. With it, one can simply use a foot pedal to step through a highlighted list of every part for a design, an invaluable visual aid when doing hand assembly.

The Pixel Pump looks great, but if you’d prefer to go the DIY route for vacuum pickup tools you would certainly be in good company. We’ve seen economical systems built for under $100, and systems built around leveraging bead-handling tools intended for hobbyists. On the extreme end there’s the minimalist approach of building a tool directly around a small electric vacuum pump.

RFID Emulator + E-paper Badge Can Be Programmed With Sound

July 22, 2023 by 9 Comments

In a way, an e-paper display makes an excellent foundation for a reprogrammable RFID card. The display only needs power during a refresh, and 125 kHz RFID tags are passive in the sense that the power for the RFID transaction comes from the reader itself. [Georgi Gerganov] has put those together in the GGtag, an open-source project for a 3.52″ e-paper badge with a trick or two up its sleeve.

One clever function is that it is programmable with sound, a feature built off another project of [Georgi]’s called ggwave, a data-to-sound (and vice-versa) framework that has been ported to just about every hardware platform one cares to imagine — including mobile phones — and can reliably send data through the air.

Transmitting data over sound is limited in throughput but has a number of advantages, not least of which is the huge range of compatible devices. There’s a web-based tool for programming the GGtag with sound available at ggtag.io that will give you a preview and let you hear how it works. The data encoding method gives transmissions a charming beep-boop quality that’s a bit reminiscent of an analog modem handshake. GGtag can also be programmed over USB serial, a faster (but somewhat less exciting) option.

The project’s GitHub repository contains GGtag’s code and technical details, and the CrowdSupply project is in the works for anyone who would prefer to buy one once they become available.

A streamlined black boiler with a headlight at the tip dwarfs the 5th wheel trailer and secondary trailer it is attached to.

Bringing A Steam Train Back From Extinction

October 11, 2022 by 74 Comments

There’s no denying that while railroads have switched to diesel and electric as their primary power sources, there’s a certain allure to the age of steam. With that in mind, a group of Pennsylvania train fans are bringing the alleged fastest steam train back from extinction.

It takes real dedication to build a 428-ton device from scratch, but these rail aficionados seem to have it in spades. Armed only with the original blueprints and a lot of passion, this team has already finished construction of the boiler and nose of the Class T1 replica which is no small feat. This puts the train at approximately 40% complete.

Some changes are planned for the locomotive including a change to fuel oil from coal and replacing the poppet valves prone to failure with camshaft-driven rotary valves. While not original hardware, these changes should make the train more reliable, and bring the world record for the fastest steam locomotive within reach. If the T1 replica can reach the 140 MPH storied of the originals, it will smash the current record of 126 MPH held by a British train, the A4 Mallard, which would be exciting indeed.

Speaking of Pennsylvania and steam, a trip to Scranton is a must for anyone interested in the age of rail.

SMT Part Counter Aims To Ease Taking Inventory

April 20, 2022 by 16 Comments

[Nick Poole] has an interesting idea for a new tool, one that has the simple goal of making accurate part counts of SMT reels as easy as pulling tape through a device. That device is the BeanCounter, an upcoming small handheld unit of his own design that counts parts as quickly as one can pull tape through a slot. The device is powered by a CR2032 cell and and works with 8 mm wide tapes up to 2 mm in height, which [Nick] says covers most 0805 or smaller sized parts, as well as things like SOT-23 transistors.

Why would one want to make such a task easier? Two compelling reasons for such a tool include: taking inventory of parts on partial reels or cut tape, and creating segments that contain a known number of parts.

The first is handy for obvious reasons, and the second is useful for things like creating kits. In fact, the usefulness of this tool for creating tape segments of fixed length is perhaps not obvious to anyone who hasn’t done it by hand. Sure, one can measure SMT tape with a ruler or a reference mark to yield a segment containing a fixed number of parts, but that involves a lot of handling and doesn’t scale up very well. In fact, the hassle of cutting tape segments accurately and repeatedly is a common pain point, so making the job easier has value.

If you looked at the photos and suspected that the big, 7-segment numeric display is done with clever PCB fabrication options (making segments by shining LEDs through PCB layers, a trick we always like to see) you’re not alone. After all, [Nick] has a lot of experience in getting clever with board fabrication, and eagle-eyed readers may even suspect that the reset and setup buttons on the edge of the tool are created by using flex PCB segments as switches. Want the nitty-gritty details? Visit the GitHub repository for the project and see it all for yourself at the CAD level.

OAK-D Depth Sensing AI Camera Gets Smaller And Lighter

October 15, 2021 by 9 Comments

The OAK-D is an open-source, full-color depth sensing camera with embedded AI capabilities, and there is now a crowdfunding campaign for a newer, lighter version called the OAK-D Lite. The new model does everything the previous one could do, combining machine vision with stereo depth sensing and an ability to run highly complex image processing tasks all on-board, freeing the host from any of the overhead involved.

Animated face with small blue dots as 3D feature markers.
An example of real-time feature tracking, now in 3D thanks to integrated depth sensing.

The OAK-D Lite camera is actually several elements together in one package: a full-color 4K camera, two greyscale cameras for stereo depth sensing, and onboard AI machine vision processing with Intel’s Movidius Myriad X processor. Tying it all together is an open-source software platform called DepthAI that wraps the camera’s functions and capabilities together into a unified whole.

The goal is to give embedded systems access to human-like visual perception in real-time, which at its core means detecting things, and identifying where they are in physical space. It does this with a combination of traditional machine vision functions (like edge detection and perspective correction), depth sensing, and the ability to plug in pre-trained convolutional neural network (CNN) models for complex tasks like object classification, pose estimation, or hand tracking in real-time.

So how is it used? Practically speaking, the OAK-D Lite is a USB device intended to be plugged into a host (running any OS), and the team has put a lot of work into making it as easy as possible. With the help of a downloadable application, the hardware can be up and running with examples in about half a minute. Integrating the device into other projects or products can be done in Python with the help of the DepthAI SDK, which provides functionality with minimal coding and configuration (and for more advanced users, there is also a full API for low-level access). Since the vision processing is all done on-board, even a Raspberry Pi Zero can be used effectively as a host.

There’s one more thing that improves the ease-of-use situation, and that’s the fact that support for the OAK-D Lite (as well as the previous OAK-D) has been added to a software suite called the Cortic Edge Platform (CEP). CEP is a block-based visual coding system that runs on a Raspberry Pi, and is aimed at anyone who wants to rapidly prototype with AI tools in a primarily visual interface, providing yet another way to glue a project together.

Earlier this year we saw the OAK-D used in a system to visually identify weeds and estimate biomass in agriculture, and it’s exciting to see a new model being released. If you’re interested, the OAK-D Lite is available at a considerable discount during the Kickstarter campaign.