Picycle Cyclist Safety

To content | To menu | To search

Monday 23 October 2017

Change is good

Time for a long awaited update for the project, and change is coming.

Web services

A new web service has been brewing in the background over the last quarter. Here's some of the key features:

  • Groups - Data and devices are now associated with Groups. This allows for controlled access to sensitive locations or data which will be shared in a controlled way.
  • Accounting - To support the above, simple user accounts will now be required to access the data. There is no charge for joining, and everyone joins the public group as standard.
  • REST API - One of the biggest features in the future release will be a user accessible API. This allows for anyone to query the Picycle database using whatever flavour of client or language they like.
  • More - Of course there is more coming in the backend, details will be available soon.

Hardware

Although not featured in this blog, the hardware has already undergone a minor revision to reduce size and cost. But there will be a huge overhaul for the upcoming release:

  • Benefits - One of the biggest bits of feedback to date has been the lack of immediate benefits for the cyclist. This will be addressed in the new release with additional functions and connectivity.
  • Proximity - Due to the impressive progress and availability of LiDAR the new hardware will ditch ultrasonic proximity and move to laser! Benefits include more options for waterproofing, improved outdoor performance and tiny form factor.
  • GPS - The old module used for GPS locating has been discontinued, a replacement has been selected which is superior and also offers support for GLONASS and Galileo.
  • Power - Power consumption has been improved by removing the requirement for a 5V supply.

Thursday 29 September 2016

First pass for MK2

MK2-A

This week the first run of our MK2 design arrived. It represents the first time that all of the processing and sensor support has been on the same board.

Watch this space for a post board population update.

Friday 19 August 2016

Whats next?

Focus has shifted slightly in the last few months. Whilst the prototype unit proved to be a valuable test for what the platform will become, it also had a couple of issues:

  • It was very power hungry. There was no sense in trying to power a full operating system (let alone a Raspberry Pi) when a simpler design would achieve the same end.
  • Size. The Zero is impressively small, but when combined with external sensors, it became quite bulky.

Therefore, the long term plan will be to open source the design for the "MK1" unit, as there is still a lot of value to be discovered in a design running a full OS (plus the ease of build and use). The back-end infrastructure will continue to be developed and supportedallowing anyone to build their own unit based off the original design to help build the concept. GitHub users will be able to commit to the source and Azure/CI will take care of the rest.

Our main focus will be to complete the development of the "MK2" design, more to come about this soon!

Monday 4 July 2016

MK1

Picycle Acquisition MK1

  • Raspberry Pi Zero
  • NEO-6M GPS Module
  • HSCR-04 Proximity Sensor


Picycle Acquisition MK1


I admit, it's not pretty or practical. But the MK1 represents a functional proof of concept for the Picycle Acquisition. The goal with this design was to put the theory into practise, to ensure the end-to-end data creation to database flow works as expected.


The MK1 also provided a platform to experiment with different types of proximity and GPS modules.


Proximity options considered:

  • Parallax Ping))) Ultrasonic Sensor - Great unit, but expensive in comparison to the HCSR-04.


GPS options considered:

  • Generic USB GPS Dongle - Good starting point, but not practical from a size perspective.
  • Adafruit Ultimate GPS Breakout - Again, a great all in one module, but expensive compared to the NEO-6M.

Friday 1 July 2016

The vision

thevision

I started this project with aim of capturing the safety conditions that cyclists encounter on the roads, with the vision to build a data set capable of guiding infrastructure planning decisions and other similar type work.
This is achieved through creating a large and constantly growing empirical data set, capturing data about the surroundings of the cyclist. The output from this used as an indicator of safety (as opposed to relying on accident statistics and perceived safety data).

Why Picycle? Because at the heart of the data capture is a Raspberry Pi!


Credit is given to Dr Rich Boakes for his support in cultivating the idea and the mentorship provided.

Not just another cycling blog

Enough blogs about people nearly getting run down or knocked off on the roads. Time to do something about it!

This blog is to discuss the work on the Picycle Cyclist safety project, and related matter...