Liikuge edasi põhisisu juurde

IoT technologies for efficient climate control

For the manufacturing industries, predictive analytics provide valuable production-related parameters. In automation and advanced manufacturing, for example, machine failures can be forecasted and predictive maintenance can be planned more precisely; in aerospace, real-time analyses can be used to monitor aircraft engines; in the automotive industry, beyond manufacturing, driver assistance technologies can be used to supply data for networking vehicles.

4 min read
© Johannes Arro

In recent years, predictive analytics has increasingly come into focus, especially as supporting technologies have made great progress such as the Internet of Things (IoT), Big Data and Machine Learning s; they are expanding the possibilities of gaining predictive insights, for example from video, audio and sensor data or log files of machines. They can be used to make forecasts, but also to establish new, so far undiscovered contexts and relationships. With this knowledge, the management of a production facility can proactively counteract possible critical situations, for example by carrying out maintenance and repairs to prevent problems.

Predictive Analytics in production help to prevent damage to machines in advance, thus avoiding production downtimes and delivery problems; overall, they increase the profitability of a company. IoT applications become particularly interesting when it comes to costly work, such as offshore operations like oil rigs or wind farms.

Predictive Analytics enable effective cost

„The challenge is always to extract information from raw data, i.e. to filter those relevant ones that provide necessary insights”, explains Henry Aljand, Business Developer at Proekspert. The Estonian company specializes in digital transformation and has achieved several European awards, including the German-Estonian Business Prize for its “Industry 4.0” solutions. Proekspert has a focus on the energy sector and climate control, but is also active in the automotive and electronics industries.

Predictive analytics are also useful for saving energy, for example in office or manufacturing buildings, especially in countries with high energy prices. Even an unnoticeable adjustment of the ambient temperature in a room or production hall can help to save several percentage points of energy and thus costs. In general, heating, ventilation, air conditioning (HVAC) systems in offices, hotels, commercial and industrial buildings or hospitals are often inefficient because weather changes, fluctuating energy costs or the thermal properties of the buildings are not taken into account.

IoT technologies for efficient climate control

In cooperation with the German shipbuilding company Rheinhold & Mahla, Proekspert developed an HVAC control solution which uses model predictive control. On two RoPax class ferries, data is gathered from vessels, simulating different control decisions and sending optimal decisions to a local PLC which controls the HVAC.

„The energy consumption of large devices such as chillers depends on the weather and internal configurations“, says Henry Aljand. „How is it possible to identify the exact energy requirements of the heating system to make it produce exactly the needed heat?“

Initially, Exploratory Data Analysis (EDA) was implemented to identify the available data and determine its quality. This also covered crew behavior, such as changing setpoints of the chillers. The second stage focused on the work process, efficiency, and potential improvement of the HVAC system. Interactive visualizations offered new viewpoints for the existing designed system and showed how the system actually works on operational conditions. This increased the efficiency and effectivity of the analysis.

In the third phase, a Correlation Analysis gave insight into how the system reacted to changes and how different parts of it affect each other. Based on the collected knowledge and observations, Proekspert created a model predictive control method. The simulation models enabled optimizations regarding the climate technology as well as HVAC savings.

Already after only a few months of operational data, the prediction models already achieve a high level of accuracy. The temperature can be predicted with an uncertainty of merely 0.2°C. The energy consumption of the HVAC systems was reduced by 10%. Control decisions (such as room temperature setpoint) are performed autonomously, i.e., not requiring human intervention. Based on predictive analytics, adaptive technology and IoT data, maintenance decisions can be supported and business knowledge gained.

„This technical solution enables exchanging data with ships even when they are at sea“, sagt Henry Aljand. „This model is also applicable to office or manufacturing buildings.“

In the mid-90s, Estonia had begun to digitalize its administrative and then its educational system. This development known as the Tiger Leap (Tiigrihüpe) has continued in business and industry. Today, Estonia is considered the most digitalized country in the world and its IoT solutions are exported to more than 120 countries around the world.

„Estonian companies are known for their IoT solutions for manufacturing industries“, says Triin Ploompuu, member of the board of the Federation of Estonian Engineering Industry. „They combine digital know-how with design and development and metalworking to create customer and application-specific mechatronic solutions at competitive prices.“

 

Original article in Produktion by Ragna Sonderleittner

See original article here

Give feedback
Give feedback