what would you want to see? i would like an update to the rules language/engine. more options and more robust.
I talked to ELK awhile back about some ideas for enhancements that I'd like to see. Apparently, to get UL listing for a security system, there are some fairly strict constraints that they have to abide by. Must not run an actual OS, must be embedded hardware, etc.
Personally, I don't care about UL certification for the security portion of it. But, pro installers that are abiding by local code are going to have to, and I assume this is the bulk of ELK's business. Buuuutttt.... maybe a solution here is to come out with a system separate from the M1 where ELK doesn't care about the UL listing, and it's targeted towards the DIY crowd.
In any case, here's what I'd like to see:
- Integrated ethernet without a separate expander
- Built in web interface for both status/arming, AND programming/setup (no ElkRP required)
- If it had a web interface, I would hope they would hire a usability expert and a real web designer to make it look slick and not be frustrating to use. Programmers are not good at this.
- AJAX for real time updating of status pages without full refresh
- RESTful API for integration with other products
- Enhanced rules support (maybe use the standard existing language for ease, but have the option of using a real programming language to perform more complex functions)
- Real OS such as linux
- Plugin capability to allow 3rd parties to write feature enhancements, drivers for other products, etc
- Database or data structure that allows a user to pull ANY current variable, setting, or status in rules or through the API
I'm sure I could think of many more items for the wishlist.
But, some food for thought... A 1.2ghz Atom board or a GuruPlug is under $100. And a USB->RS-485 interface is about $11 (I just bought one to play with). Replicating the existing logic of the M1 in software would not necessarily be a monumental undertaking for ELK,
and the RS-485 interface would allow the system to use all of the existing hardware (keypads, I/O boards, etc). They have built in sound. They have network. Running linux on the device would allow them to add decent text to speech, and possibly voice recognition (MisterHouse did it). FreeSwitch could be installed on it for allowing it to dial out via a SIP trunk, or allow one to dial in and get an IVR menu. An Atom board would even have enough power to be a phone switch for the whole house and people could purchase VoIP phones.
By building a basic system on an OS that replicates the existing functionality of the M1, but allowing plugins and open access to it, ELK can allow the community to expand the functionality of the system via plugins or modifications without expending a huge amount of internal resources to add all of the features that people want. Their software could be written in a compiled language and require a serial number to prevent piracy, but plugins and automation scripts could easily be written in whatever language the user felt comfortable with (I'm a big ruby fan, even though I'm a lot more proficient in Perl)
I realize that there are constraints on what can be used when trying to achieve a UL listing for security... But maybe that shouldn't be the primary focus of a new product. Obviously having that feature there would be a requirement for most DIY guys considering purchasing a system. Personally, I think the UL's requirements need to be updated, they are holding companies back from innovating, and making all of the hardware on the market use 20-30 year old technology at the core with newer tech having to be duct taped in somehow. This certainly is not a rip on Elk or HAI, or anyone else making security/HA systems. I'm just saying that they are being hobbled by the current requirements.