Jolly Good Sensor Project
What we have here is the temperatures logged from a set of dual-probe sensor units located at the Jolly Good Beer warehouse. The key motivation for this is to display the stability of the temperature of our beer stock. That’s the solid solid blue line. We believe strongly in keeping beer cold to preserve its flavour – to keep it as “fresh” as possible. Beer is a product that, in the vast majority of cases, degrades with time – and this reduction in quality is accelerated by heat. We feel that providing evidence of the work we do to protect the quality of the awesome beer we sell is important – and it is part of our commitment to achieving a full coldchain supply chain for good beer.
The darker lines are probes that are in the middle of a 30l KeyKeg of water. This is to replicate as best we can the temperature the beer would be at in these environments, whilst the paler lines in matching colour indicate the temperature of the air in the vicinity of the keg – the ambient or environmental temperature you could say.
Note how the red and orange lines fluctuate quite rapidly in line with the ambient temperatures – and warm quite quickly as the day warms. Whilst our coldstore keykeg maintains a comparatively flat line just below 4°C. At this time (end April 2018) the average difference in temperature is nearly 10C already… that could be approximately equivalent to a 3x reduction in shelf-life according the Dr Charles Bamforth in “Freshness“. Or, another way of putting it is our methods even in spring are granting the beer 3x extended freshness versus standard ambient storage of keg and smallpack in the UK. This will be the case for at least 9 months of the year, and even for most of the other three months our 4°C storage will be better than ambient – and stable.
Also – we’d like to extend a completely open invitation to breweries (especially those supplying us), industry folk, beer writers/journalists, and just “beer people” in general to pop by (preferably with notice, but hey rock up by surprise if you don’t trust us) to see for themselves how we do things. There is no “smoke and mirrors” here.
We plan to deploy more of these sensors – at least two more in different parts of the coldstore (the parts and cabling have been bought for these) and we have a plan to do some useful temperature change measurement between different start points to reach equilibrium temperature different environmental temperatures. I’ve some kit to do remote logging too so we can do in-transit temperatures, and extending on this deploy monitoring nodes to customers who care about beer like we do. So much to do… so little time.
This has been implemented using Sonoff TH10 units that have had two DS18B20 probes connected – they have been reprogrammed with the Tasmota firmware. The units send out an MQTT message every 60 seconds with the temperature which a small Python daemon is subscribed to and logs to a MySQL database, a separate Python daemon regularly generates a JSON format snapshot of this data and uploads it to the webserver – where it can be viewed in graph form on this page. Note that the accuracy of these sensors is give or take 0.5 of a degree, as per the DS18B20 datasheet. Just ask and I can share the code/details of this – which is a bit of a “rough hack” really.