Can it really be true, that no-one, on a planet of almost 8 billion people, has successfully built a business around creating open security cameras? Even with very basic requirements? In this article, I outline the security camera project’s needs and ponder on the state of business.
This post is part of a series
- The open security camera project
- MVP’s for the camera and existing alternatives
- Scoping hardware options
- Creating the case
- Tackling mounting
- Assembly and installation
- Software stack
- Security camera hardware or as a service from Molnix
- The how-to for DIY people
Requirements (MVP’s) for the solution
The first and most critical aspect of the solution is the level of trust a camera system has to be able to deliver. There are two ways of achieving this with us: through openness and transparency or through earned trust such as reputability and established relationships. At Molnix, we always start at the first alternative for technical solutions. This the approach we adopt in this article series.
The following are the minimum set of criteria that the solution has to fulfill:
- Fully open source software stack, firmware exemptions possible
- Sufficiently open hardware stack (i.e. network path is auditable)
- Night vision capable
- Indoor and outdoor
- Single cable installation
- Networked IP video stream
- OpenHAB compatible
This paragraph could easily become much too long. So instead of evaluating every solution I had a look at, here’s what I’ve done. I took two approaches as my starting point: hardware and relevant software projects. Then, I screened them for openness and features.
Starting from hardware, I was unable to find one that would come with either as a hardware-only setup or with an option to flash an open source software stack. It seems all of the cameras out there come designed for specific systems or are stand-alone. Some have custom ROMs available, but again, I was unable to find an open source custom ROM. The closest I would get are some blob-less webcams and the Raspberry Pi camera variants.
The next approach I tried was to look at the major open source security camera software projects. I was hoping to find a reference implementation or even developer kit. However, these did either not exist or did not cover the required feature set. Most of them support Raspberry Pi, though, which is promising.
As it looks at this stage, a Raspberry Pi-based hardware platform is most promising. It also leaves the options open for multiple software solutions.
The discussion is open: did I miss something? Is there a camera that I can buy out there? Do (did) I really have to design our own?
In the next post, I describe the hardware we ended up selecting for the project.