Water & Wastewater Treatment

34 | FEBRUARY 2015 | WWT | www.wwtonline.co.uk In the know Technically speaking: connected water management and machine condition monitor- ing require sophisticated, accurate multi-sensor systems and associated signal processing. • Costs have been driven down dramatically. Applications, such as personal health monitors, solid waste bins and most water sensors have no electrical power available - they have to operate for long periods of time on a small battery. • Devices in the fi eld are becoming smarter, and will be based on off - the-shelf, low-cost hardware. They contain intelligence that is specifi c to the function that the Thing has to provide. For example, a pump station controller will be equipped with diagnostics for monitoring the condi- tion of the pumps, their hydraulic performance and energy effi ciency, and will be able to call for a main- tenance intervention when required (Aquamatix). A pressure sensor with embedded so ware which can detect and ana- lyse transients and fl ow data; clever signal processing algorithms which extract the information value from the sensor and send critical events to a dynamic hydraulic model or decision- support system (Infrasense Labs). A wireless ultrasonic level trans- mitter with a ten-year battery, which will transmit sludge tank levels when there is a signifi cant change in con- tents and report solids inventory to a 'travelling salesman' optimal route planner (Enevo). These are a few ex- amples of devices available today, and the rate of innovation will accelerate. Open, wireless communication networks A multi-sensor application consisting of many sensors of a similar type, such as pressure or water quality, distrib- uted across a wide geographic area - or a set of multi-parameter sensors in a personal health monitor or monitoring a water body - need to be intercon- nected with other machines which use the data. Communications networks, in most cases, will be wireless. Data transfer protocols such as Bluetooth (low-power, short-range) WiFi & Zigbee (medium –range) and GSM (medium-power, wide-area coverage) are becoming pervasive and low-cost. Security is a concern but is designed into the protocols and with multi-layer authentication at the edge device. Open standards are important to ensure multi-vendor interoper- ability – a term describing products from diff erent manufacturers working together. Point-based protocols such as DNP3, which were popular in 1990s for telemetry, are being replaced by message-based protocols that transfer whole blocks of pre-processed data from the intelligent edge device. Most open protocols will be based on IP (internet protocol), however some local networks, especially if ultra low power is a requirement, as it is in water, will not use IP because it is not energy-effi cient. Various gateways will be necessary to channel data from legacy and IoT devices via a public or private internet to data processing facilities. Cloud computing Cloud computing has drastically reduced the cost of data processing and storage due to the low cost of ownership and the lack of a need for server infrastructure or the additional cost of a data centre. A new term 'FOG computing' (not fats oils & greases) describes a more local distributed cloud resource which brings the processing to the data rather than pushing the data to the cloud. As the edge devices become smarter, the data that needs to be transmitted as messages will be less frequent and higher value. The current perception of "drowning in data" will change to become the right data in the right con- text at the right time – a pipedream perhaps but the potential is certainly there. Big data analytics Big data analytics and machine learn- ing, a vast array of techniques for ex- tracting meaning and understanding from data, data-mining, self-learning algorithms and model-based reason- ing, are a few of the mechanisms used for building automated deci- sion support systems (DSS). When coupled with advanced data visu- alisation, actionable insight can be gained from real-time data coupled with new and legacy mathematical models. Integration with GIS and asset management and other enter- prise business applications, such as energy optimisation, will also be critical business capabilities. Mobile computing Smartphones and tablets deliver actionable insight to human users on the operational front-line. Mobile browsers and apps also provide a rich platform for user interaction and data collection, and become data processing Things in their own right. So how will all this new technol- ogy impact on water management? Agriculture accounts for around 70% of the world's fresh water resources; there are already many smart systems in service for precision ir- rigation and crop yield management, which signifi cantly reduce water demand. Aquamatix, a specialist technology company which is devel- oping IoT for the water industry, is designing a system for a 220 year old canal which will take a weather feed from a crowd-sourced web service and use the data to predict rainfall; this will allow optimisation of stor- Almost every aspect of water management could be plugged into the Internet of Things (IoT)

• Cover