I’m David Rowe, an electronic engineer living in Adelaide, South Australia. My mission is to improve the world – just a little bit. I do this through developing innovative hardware and software for telephony. This hardware and software can help many people, for example through improved, low cost radio communication for poor people in remote areas of the world. So rather than making money from his technology, I give it away.
In January 2006 I quit corporate life to develop open telephony hardware and software full time. In most of these projects I have been part of a community of like minded, open source developers. Open source is one of the reasons why the technology works so well.
Rowetel is the name of the business I operate from a home office in Adelaide, South Australia. For income I sell telephony hardware and engage in contract HF modem and speech compression work. I use the income from my business to fund my work on open software and hardware projects.
Here are the projects I’m working on today.
Projects I have worked on include the SM1000 Digital Voice adapter for HF Radio, a sub 2000 bit/s speech codec called Codec 2, PC software for Digital Voice over HF Radio called FreeDV, the Oslec open source line echo canceller, the open hardware IP0X range of embedded Asterisk IP-PBXes, and the Mesh Potato – an 802.11bg mesh router with telephony. A complete list is on my projects page.
If you think my work is worthwhile, then please consider donating your time, skills, equipment, or a few dollars.
I have reached the point in my life where I have enough money to be comfortable, and would now like to improve the world a little. Rather than use my technical skills to accumulate more wealth, I would like to help people. So I keep my expenses low, live on a modest income, and spend most of my time working on open source projects. I choose to give the technology away to anyone who finds it useful. Craig Newmark’s nerd values.
- I write a popular blog that attracts over 150,000 visits (unique IPs) a month. I publish a detailed post about once a week and use a narrative (story telling) style to talk about the projects I am working on. In particular I like to talk about what went wrong and how we fixed it. It’s really important for me to teach people what I have learnt, if I fix an interesting bug I just have to blog on it. I really like the idea of leaving a trail of breadcrumbs on the Internet. Teaching my readers something new with each post is one way I can improve the world a little.
- The Codec 2 project started in September 2009. This is an open source codec designed for speech compression at 2400 bit/s and below. Incumbent closed source 2400 bit/s codecs require expensive and awkward licenses and are stifling innovation. Applications include VOIP trunking, digital voice over HF and VHF radio, e.g. Amateur Radio, developing world and remote area communications, military, police and emergency services. This project has also spun off a HF FDM modem, FreeDV, and the SM1000.
- FreeDV is a GUI application for Windows and Linux that allows any SSB radio to be used for low bit rate digital voice. Compared to traditional HF SSB it offers speech free from channel noise and interference, and uses just 1100Hz of bandwidth. It packages Codec 2 and the FDMDV modem into an easy to use, end user application. FreeDV was built by an international team of Radio Amateurs working together on coding, design, user interface and testing. There are several Digital Voice systems for HF and VHF radio, but FreeDV is unique as it uses 100% Free Software. It represents an open source path for 21st century Amateur Radio, rather than a future locked into a single manufacturers closed technology. For me it is also a way to experiment with low bandwidth digital speech systems which have applications such as “last 100 mile” links in the developing world.
- The SM1000 is a small box in a microphone form factor that lets you use FreeDV without a PC. It can convert any analog HF or VHF radio to Digital Voice.
- The Village Telco is a DIY telephone company that uses mesh Wifi and VOIP to build telephone networks without infrastructure like cell phone towers or land lines. The goal of the Village Telco is to provide affordable telephony for people in the developing world. In June 2008 I attended the inaugural Village Telco workshop. We needed mesh networking hardware with telephony but off the shelf hardware was not good enough. So we took the radical decision to develop and put into volume production a full custom open hardware product designed specifically for our needs – the Mesh Potato. From 2008 ato 2010 I managed the Mesh Potato project, lead the design and development team, liaised with the manufacturer (Atcom), and worked on many of the tough engineering challenges of the project. This project was supported by the Shuttleworth Foundation. More information on my Village Telco and Mesh Potato work here.
- In March 2010 the Dili Village Telco project started, the world’s first roll out of Village Telco technology. Over 18 months we deployed 60 Mesh Potatoes to implement 3 free local call telephone networks in Timor Leste. I managed the project, trained the Timorese team, and troubleshooted technical problems. The project gathered important technical, social, and business model data for the Village Telco. This project was supported by ISIF and ISOC grants and Atcom, who kindly manufactured a special batch of Mesh Potatoes for the project.
- A high performance, open source echo canceller (Oslec) was released in June 2007. Oslec runs on both x86 and Blackfin platforms, and provides high quality, free echo cancellation to thousands of people around the world. Oslec solved a major, long term problem with Asterisk – poor echo cancellation. Until Oslec was developed, Asterisk users were forced to pay for license fees or buy expensive hardware for effective echo cancellation. Oslec fixed this problem overnight and echo simply disappeared for many people in the Asterisk community.
- In September 2005 I ported Asterisk to the Blackfin CPU. This opened the way for low cost, low power, but powerful Asterisk IP-PBXes that don’t need a PC. In December 2005 the radical decision to develop open telephony hardware was made and the Free Telephony Project started.
- The IP04 Four Port IP-PBX was developed in early 2007. The first phone call was made in April 2007 and with the kind help of Atcom production IP04s were released in July 2007 . The IP04 is unique – a production IP-PBX that is open hardware. It combined community-developed hardware and software to produce a low cost, completely open IP-PBX. The IP04 is now a stable, proven design with thousands of units in use all over the world. Many companies have adopted the IP04 design as the engine for their internal IP-PBX product development.
- The $10 ATA is a low cost, open hardware, FXS port design. It uses an Atmel micro-controller to implement both audio I/O and a DC-DC converter for telephony battery and ring voltages. It can be assembled from easy to find analog parts or even e-waste. It is being developed to lower the production cost of the Mesh Potato, and can also be used to turn any OpenWRT router into a mesh telephony device.