20141003: Some brief notes from IoT Boost in Liverpool. See: https://cde.catapult.org.uk/blog/-/asset_publisher/Ci2GCCoQyuZS/blog/a-fantastic-day-at-internet-of-things-boost-in-liverpo-1 I went with a "tell me what I don't know that I don't know" open mind, and I got a few pleasant unexpected insights... Security / Paul Galwas: Amongst other nuggets Paul focussed my thinking on minimising the amount of our software (and hardware) that has to be right (and validated) to protect sensitive data and control, whether for authentication or encryption. Basically minimising the attack surface and the amount of stuff we have to that we have to sweat over. Selectively protecting part of particular payloads rather than the entire data channel takes most of the radio hardware out of scope for example. And given that we're mainly talking about protecting in-flight data (not stored to disc for example) and for a limited radius (and time) AES128 is probably enough for now rather than AES256 for example. Buzz-phrases to search on for more: "Protection profile" "Common Criteria" (ISO/IEC 15408) "Target of Evaluation" Also Paul suggested that we actually publish our security evaluation as a USP. Marketing / Anjali Ramachandran: Anjali helped me think how we might have to segment our market and branding for example between FOSSH, social, retail, and commercial deployments, and how we might do some relatively cheap market research ranging from SurveyMonkey to focus groups. We need to construct brand guidelines for each segment/brand. Also, as others stressed to me, design aesthetic is important; even if we commoditise some products to make them affordable they should still look 'designed' and 'premium', especially for retail/consumer. Manufacturing / Josh Valman: Very interesting discussion including some on branding, in particular that 'OpenTRV' is probably too much of a mouthful for a compatibility logo (like 'WiFi'). The main message was that getting any consumer product from design stage through manufacturing to retail shelves or the equivalent takes at least 6 months, which is exactly in line with what I have heard elsewhere. I toyed with a few other 'eco' product ideas around the OpenTRV / DD area, and Josh kindly offered to look them over for sanity... (At lunch I also actually got to meet OpenTRV's radio guru Mike Stirling in person for the first time ever, which was great...) Wireless Connectivity / Hamid Falaki: Many good thoughts including putting some interesting ideas into my head about satellite comms, eg with Avanti, and about use of spread-spectrum/UWB for robustness if nothing else. Embedded Software Development / Adrian McEwen: This became an an interesting chat about the Arduino GSM shield, variants of it and its library, and the Quectel M10 modem chip. Oh dear, I massively overran the allocated time. The Teensy (3.0/ARM?) was suggested as another possible OpenTRV dev platform. Thanks to Designswarm (Alexandra and Ana) for making good stuff happen...