Smart buildings influencing the living environment Neeme Takis

Smart buildings influencing the living environment Neeme Takis Rakvere Smart House Competence Centre Uniflex Systems, ITvilla -> Base. N Baltics neeme@takis. ee

Content • What's important inside a smart building? • Which techniques to use in – improving occupants' comfort, – achieving energy savings, – avoiding damages, – increasing security • What are the known problems and likely development paths?

What’s important? • A smart building should – improve occupants' comfort, – achieve energy savings where possible, – avoid or reduce damages by sending warnings – increase security compared to the traditional. • How? – Using the cheapening power of communication and computing we have at our reach today – Forcing the subsystems to work together, not independently (or even against each other)

Improving comfort. Control Quality • Get rid of open loop – Heating example: most of the installed water based heating systems are primitive, setting the on-flow temperature according to the current outdoor temperature (not interested in the actual result at all)

Improving comfort. Control Quality – If the actual result is not measured by the control system, then no disturbances from the sun radiation or wind can be compensated – Room thermostats are common to take actual temperature into account, but due to thermal inertia the result of their action is a fluctuation around the setpoint temperature

Improving comfort. Control Quality – The solution: closed loop control • Single loop for simple cases

Improving comfort. Control Quality • Cascaded loops for real life (with limit settings)

Improving comfort. Control Quality • Use predictive information – Weather forecast can be used well before the significant change in the outdoor environment conditions, for time shifted • change in the setpoints • heating stopping before the outdoor temperature increase – The amount of time shifting depends • On change direction • On building parameters (heat losses, installed power of heating) • On “step height” of the outdoor temperature drop or increase

Achieve savings. General • In some cases improving the control quality will already result in energy savings as well (by avoiding unnecessary heating) • Other things to do – Eliminate the unnecessary energy spendings • like simultaneous heating and cooling… – Use the hourly tariff and time-shift the electricity consumption peaks if possible • washing the dishes in the cheapest hours • storing the heating energy into a tank wisely • Switch off heat pump during peak hours if possible

Achieve savings. Dishwasher example • There are no diswashers (to my knowledge) that can find the cheapest hour on their own • But they all continue after the power break from where they were • So start the washer and press a button (actual or virtual) to stop the process • Power will be returned on time to use the cheapest possible time to be ready with the washing before the next morning

Achieve savings. Heat pump example • With air-to-air heat pumps the efficiency of the pump depends on the outdoor temperature (may even drop below 1 during cold weather!) • Depending on the availability of other sources of energy (gas, wood, direct electric heating) it may be reasonable to switch between them depending on the hourly prices on electric energy • …Or just switch off the heat pump during the peak hours.

Achieve savings. Direct heating example • If a storage tank is in use for nightly heating, replace the primitive day/night tariff switching or simple timer based control with a price-aware solution • Estimating how much heat energy is needed next day, charge the tank with heat energy during the cheapest hours • Charge to the needed level only (not more)

Avoiding damages. Water leakage example • Water leaks are costly and hard to detect – Water can make more harm than fire – Water detectors cannot be installed everywhere • A water meter with good (1 liter) resolution will help, if (any of the following) – the water consumption pattern is known – building occupancy is know – movements in the rooms can be detected • Some decision-making power is needed. Doable.

Avoiding damages. Equipment monitoring • The increased relative to the historical average (and working conditions) energy consumption of one or another device may be a signal of – Improperly closed refrigerator door – Loss of cooling agent in the cooling system – Decreased efficiency of pumps or filters • A professional can see the abnormalities and raise warnings by looking at the trending graphs • The building automation system should do the same based on historical and current data

Improving (access) security • The traditional approach – armed or not • A better approach - always on watch – Analyze the movement and door/window position sensor data - not independently as before, but by combining and comparing with signals from the other sensors – Instead of the primitive zone violation signal the legal and illegal situations and/or movement tracks can be detected in real time

How? The main problem: interoperability and fragmentation (1) • Lack of systems interoperability standards – BACnet, Lon. Works, KNX or other aging protocols are not well suited for interoperability – Several other protocols competing • Current M 2 M efforts are fragmented – Competing technologies promoted – No reuse other than basic networking • Certain conflict between the traditional and Io. T automation way of thinking – Including local vs centralized vs distributed approach in functionality, data storage, decision-making, configuration, optimization, security

How? The main problem: interoperability and fragmentation (2) • Traditional approach: separate systems – – – – Access control Fire alarm system Computer network Phone network Lighting control network Utility meters network Building automation network (if exists) • Ventilation subsystem • Heating subsystem • Cooling sybsystem • Each of the systems – need special knowledge – will fail without maintenance sooner or later

How? Using the power of communication and computing • Io. T-styled approach – Distributed I/O (input/output interfacing) • • Any sensor as the information source for inputs Any actuator as the control target by outputs All connected to the closest I/O device Forming a “cloud” inside the building – Control and monitoring application(s) • Using inputs and generating the outputs • Using data from external servers • Feeding data to extrernal servers (incl monitoring!)

How? . . . working together • Possible (over the subsystems) within both traditional and Io. T-styled approach – Traditional: add an integration gateway to talk to all subsystems (in their preferred language) • Case per case custom made software • Limitations from the subsystems to account – Io. T styled: the software has access to anything, can do anything • Easy integration, data reusing, even self-learning • Risk to affect everything by a single software bug! • Data/comm security to be taken to a new level (compared to the traditional systems)

Io. T will change the game • Everything is connected • The challenges – get the data in an interoperable format – break the vertical data silos – harvest the data across domains • Consumer Io. T – mostly wireless – also has to deal with the wired legacy of KNX, Mbus, DALI, etc • Industrial Io. T – mostly wired – has to deal with the legacy using specialized protocols like Lon. Works, Device. Net, Profi. Bus, CAN, Modbus, etc.

Bus vs Broker • There are two dominant architectures for data exchange protocols: – Bus-based • Messages to be delivered directly to the receiver – Broker-based • The broker controls the distribution of the information (including storing, forwarding, filtering, prioritizing) • The client may switch between the roles of publisher and subscriber

Message vs Data - centric information distribution • Message centric – focus on the delivery of each message, regardless of the payload data – will likely force to process all skipped messages (with already outdated content) after a communication break • Data-centric – Focus on delivering the meaningful / currently valid data instead of all messages – Updates the known content with the last values only after a communication break (possible history loss)

Some protocols to be present in the future Bus-based Broker-based Messagecentric REST MQTT, JMS, AMQP Data-centric DDS, XMPP Co. AP

Choices for the consumer • Select, buy and connect the components – the outcome is quite limited unfortunately • DIY approach (based on Arduino, Raspberry Pi, …) – Takes time, some things are great and/or unique, but many things never get ready – Nice hobby, but the outcome is not a product to sell • a lot more than a working prototype is needed for business • Industrial installation – The smart properties installed by the professionals, possibly within the building process – Similarities with the car industry…

Offerings from related companies • Base. N Corporation OY (www. basen. net) – Big Data monitoring system, 800 k transactions per second received and stored currently. Data network and utility measurements (latter for dwelling houses). • Uniflex Systems OÜ (www. uniflex. ee) – Uni. SCADA as a Web-based monitoring and remote maintenance system, able to replace the traditional SCADA systems, compatible with well-known ITmonitoring tool called Nagios • ITvilla OÜ (www. itvilla. ee), now Basen Baltics OÜ – Integrator / solution provider for automation systems – Manufacturer of industrial Linux-controllers and flexible I/O modules

Thank you for your attention! • neeme@takis. ee, +372 5010066