Artificial intelligence (A.I.), garnered its highest worldwide attentions when AlphaGo, an A.I. program developed by DeepMind, relentlessly defeated one of the best board game Go players in 2016. In recently years, A.I. has been applied to many fields of industry and becoming one of the hottest pioneer technologies for our future. Among these find the optimization of chiller plant system.
As an expert in energy saving solutions for Chiller system, BOCA never falls behind.
While having a handsome setup and infrastructure with brilliant energy saving capacity in place is challenging, manage, maintain and even further optimize the system in daily operation is even more. For this aim BOCA is developing the most effective tools based on the different types of A.I.
Rule based A.I. system embeds list of fixed rules or formulas derived from summarization of expert knowledge and experience. These rules form a model which will govern the operation and maintenance of the whole chiller system. They are always generalized and not specific enough for customizing a particular system. Because the rules must be clearly defined, as a whole they are hard to modify and too simple or rigid for a large and complex system. Moreover, the rules are set according to initial conditions and factory parameters of the system, they are not able to fine tune for changing situation as ageing problems rendering along the life cycle of system elements.
BocaSmart A.I. takes whichever type of A.I. works best. Machine learning though stands head and shoulders above the capabilities of rule based, it would not be wise to exclude the latter. Machine learning models take a lot of time to acclimate and learn from big data, which are not possible to be collected at the initial stage of running a system, yet the training time is very long. While machine learning’s strength is in the amount of data it can analyze and track in real time, a human’s strength is in providing context and intuition and analyzing edge-case scenarios and generalizing into basic rules. BOCA A.I. employs operating rules and machine learning in tandem to achieve best possible performance.
Machine Learning A.I. system uses the big data collected by sensors in site to model, predict and diagnosis a HVAC system to optimize energy consumption. It contains no rules nor formulas but only data and its genetic Algorithm to generate solutions and calculate the outcomes. This process always involves the use of an Artificial Neural Network (ANN) to simulate the thinking, decision making and learning function of human being. It is capable of dealing with complex system as it takes little concerns over variables and their inter-dependency. Because it takes latest data collected online from the site into calculation, it always reflects the latest conditions of the system.
The development of the BocaSmart A.I. is undergoing and will not stop taking its steps ahead as technology never cease advancing. A new era is coming for new technology such as 5G, internet of Things (IoT), A.I. breakthrough as well as hardware improvement from WIFI sensors, nested thermostat to A.I. pumps. We will keep our A.I. ever-evolving.
With more and more sensors and the use of latest 5G and IoT technology, now we can collect more online data, from system data such as entering water temperature, leaving water temperature or water flow rate to operating ambient condition data. we are storing and analyzing a big data pool in order to build an efficient and accurate ANN structure.
The Algorithm utilizes historical and latest data to compute, determine and apply weights on neuron factors for decision making and control. It covers the diagnosis and prognosis of different areas of a system, from sequence control to temperature control for chiller plants, cooling towers, water supply and pressure control. We are shaping the core body.
A software will embody the core algorithm and pack it into a control console platform which integrates all of the IoT data and control modules and present on a central dashboard for convenient daily monitor and operation. This software can be installed on the usage side or the maintenance side or merged into the Building Management System (BMS) on the client side.
Thomas Edison’s quote always keep us moving forward. Tested three thousands materials to invent his light bulb, Edison demonstrated the advance of technology are just full of hard toil.
We will never satisfy with what we have achieved, our R&D talents are now developing the automatic optimization operation program and console powered by Artificial Intelligence. In our theoretical calculation, there is a room of another 10% or more energy saving advancement within the same setting and infrastructure up for grabs.
We are going ahead full throttle on this so expect the even higher energy saving from BOCA in future!