Sustainability

Placeholder - change this Minxing

We offer data analytics, predictive emissions modeling, and air quality monitoring via internet of things (IoT) devices with machine learning (ML) algorithms.

OpenPEMS™

OpenPEMS™ is a free and open source Predictive Emissions Monitoring System (PEMS) that trains models to predictive air emissions, such as NOx and SO2, which are typically monitored by Continuous Emissions Monitoring Systems (CEMSs) OpenPEMS™ is inspired by OpenAI's ethos of accessibility and empowerment. Our vision is to make the Artificial Intelligent (AI) and Machine Learning (ML) technologies accessible to a wider industrial audience, reducing costs associated with air emissions monitoring.

The roots of PEMSs development trace back to my PhD study conducted in collaboration with industry partner Cenovus Energy. The goal is to provide Alberta industry with AI-based tools as an alternative to their CEMSs. The goal led to the IEEE publication in 2019 with Cenovus Energy as one of the co-authors. The model developed by Cenovus and I was ultimately approved by Alberta Energy Regulator and Alberta Environment and Protected Area (APEA) for regulatory reporting in 2022. Since then, a number of Calgary based companies tried to copycat the success. Check out who they are here.

Learn more about this project.

OpenPEMS™ is perfect for companies that looks for CEMS backup systems or alternatives to CEMSs. For facilities that look for integrated PEMS solutions, key players such as ABB, Rockwell, and Honeywell provide comprehensive PEMS offerings that seamlessly integrate with Distributed Control Systems (DCS), data storage systems, and reporting systems.

In the landscape of PEMSs, long-standing players like CMC Solutions have been delivering PEMS solutions for over a decade and installed more than 100 systems for regulatory reporting.

Data Analytics

We provide data analytical services to discover useful information from datasets. Below is an examples of our data analytical projects:

Fuel consumption analysis and cap and trade system evaluation for Canadian in situ oil sands extraction

In this project, we retrieved operating fuel use data from a public database for 18 in situ oil sands extraction schemes. From 2015 to 2019, the weighted average of fuel use was 0.23 103 m3/m3 undiluted bitumen. The weighted averages of fuel use for the schemes using Steam Assisted Gravity Drainage (SAGD) and Cyclic Steam Stimulation (CSS) were 0.20 103 m3 fuel/m3 of undiluted bitumen and 0.34 103 m3 fuel/m3 of undiluted bitumen, respectively. The average EIs for SAGD ranged from 0.25 metric ton (t) CO2e/m3 to 0.98 t CO2e/m3, and the average EIs for CSS ranged from 0.61 t CO2e/m3 to 1.18 t CO2e/m3. In addition, the study pointed out that production ramping up and maturity of reservoirs contributed to the decline in EIs.

Learn more about this project.

Internet of Things

We have experience with remote sensing and internet of things. Below is an example of our IoT project:

Evaluation and calibration of a low-cost particle sensor in ambient conditions using machine-learning methods

The project evaluated the performance of this low-cost PM sensor in ambient conditions and calibrated its readings using simple linear regression (SLR), multiple linear regression (MLR), and two more powerful machine-learning algorithms using random search techniques for the best model architectures.

The two machine-learning algorithms are XGBoost and a feedforward neural network (NN). Field evaluation showed that the Pearson correlation (r) between the low-cost sensor and the SHARP instrument was 0.78. The Fligner and Killeen (F–K) test indicated a statistically significant difference between the variances of the PM2.5 values by the low-cost sensor and the SHARP instrument. Large overestimations by the low-cost sensor before calibration were observed in the field and were believed to be caused by the variation of ambient relative humidity. The root mean square error (RMSE) was 9.93 when comparing the low-cost sensor with the SHARP instrument. The calibration by the feedforward NN had the smallest RMSE of 3.91 in the test dataset compared to the calibrations by SLR (4.91), MLR (4.65), and XGBoost (4.19). After calibrations, the F–K test using the test dataset showed that the variances of the PM2.5 values by the NN, XGBoost, and the reference method were not statistically significantly different.

Learn more about this project.