For our first project log, I'll start with telling you our story - how we came to work on this project and some details about our background.

Tom and I met in grad school. Tom was doing an MEng in computer engineering and was inventing some kind of electromechanical valve train for cars. I had just finished my MSc in wireless mesh networks and was about to start a PhD in heterogeneous wireless networks (think of using all of the wireless there is - bt, wifi, 4g, etc. in combination or some subset of them). Tom and I both ta-ed a course together and started talking about robotics and wireless and how we could build something awesome together.

A little while later, Tom started working at a company (while still doing his masters) doing spectroscopy and robotics. One of the projects he was involved in was a Mars rover prototype for the Canadian Space Agency. It was here he started to notice how often the wheel is re-invented during robotics. Adding sensors, motors and components is the same work being done all over the world at similar companies. There was always complex circuits, specialized code and device drivers and customization. This is where we knew we could start to make a difference. Tom was also motivated by some previous work at a water engineering company which used Programmable Logic Controllers (PLCs). In the 1970s, the PLC was created so that people didn't need to re-invent the wheel in factories when connecting up machines.

After a bit of preliminary work in this direction, we also started to think about how I could bring my expertise to the mix. We knew that moving robots were going to be the future. With my research background and work on wireless network protocols, we knew that one of the challenges when things move is keeping them connected. We decided to start combining this ability to remain connected to other system to the idea of easily connecting sensors and parts. We created a software layer that made networks automatically organize themselves so that groups of robots could communicate easily.

Also motivated by the PLC, we wanted our robots to be focused on data. Increasingly we are now finding that this is important. Robots are often created to monitor environments where it is unsafe or impossible for humans. This means lots of sensors and lots of data. For this we added the ability to easily get sensor data into algorithms (for autonomy and for inter-operability with ROS, OpenCV and other common robot tools), and the ability to easily send this data to the Internet. We have created a simple cloud service on Amazon Web Services (AWS) that lets you see all of the robots in your fleet, monitor and visualize their data in real-time.

At the same time we were developing this, Tom was completing his PhD in computer engineering working on a food processing machine that trimmed the fat off of pork loins more efficiently and I worked on my PhD in wireless communications specializing in heterogeneous wireless networks. I also got involved in projects involving cognitive agents, machine learning and other things useful for robotics.

In the upcoming logs, we'll show off some of the features that work now in more detail, and describe the direction we'd like to see this project go in the future.