Data Science cases

In a new university hospital currently under construction in Canada, all goods such as food, medication, linens, and waste will be transported to their destinations using AGVs (Automated Guided Vehicles). CQM is supplying the systems and the control software. This operation involves a high volume of movements and routing decisions for the AGVs (VOLUME). An added challenge is that the AGVs must transport goods between different floors across multiple hospital towers. CQM was asked to develop an algorithm for the most efficient elevator control system across these three towers. Efficient flow is crucial for making smart use of available elevator capacity. (Prescriptive) Additionally, high-priority AGVs (e.g., those carrying hot meals) must be given precedence. Through simulation, CQM provided insight into the duration of various process steps and identified which timings could be influenced by a smart algorithm (Descriptive). Furthermore, this algorithm is required to deliver an optimal solution within a maximum of 0.5 seconds (VELOCITY).

AgroEnergy is the leading energy specialist for the agricultural sector in the Netherlands. The company helps growers achieve optimal results in their energy trading. Last year, AgroEnergy introduced BiedOptimaal, an innovative plug-in that simplifies the growers' bidding process specifically, determining the price and quantity for purchasing gas and electricity. BiedOptimaal calculates APX bids that minimize the grower's heat production costs. CQM assisted AgroEnergy in developing BiedOptimaal, utilizing a data science approach. BiedOptimaal runs daily, just before the APX bids must be submitted. It utilizes a wide variety of real-time data sources. The system constantly uses the latest data such as weather forecasts and synchronized buffer fill levels to rapidly generate bidding recommendations. Forecasts regarding heat demand and energy prices (Predictive) are integrated into an optimization model that determines the optimal bid for the grower (Prescriptive). BiedOptimaal has been available to growers since October 2014, and during 2015, it is being adapted for use by growers who use supplemental lighting. AgroEnergy aims to offer a similar service to other sectors, such as the built environment.

As early as 2008, Corus (now Tata) engaged CQM for a data science project aimed at reducing customer complaints regarding steel quality. The project drew upon a wide variety of data sources within Corus. By placing a strong emphasis on the objective, it became possible to clearly distinguish relevant data. To unlock the project's potential, large volumes of data had to be linked together; however, the data existed in various formats, making straightforward integration difficult. Furthermore, the data contained hard-to-detect errors or inconsistencies (veracity issues). There were also practical challenges in tracking steel coils through the production process. Steel coils undergo repeated uncoiling and recoiling, causing the sequence of the material to be reversed each time. Additionally, as the coils are rolled thinner, their length increases, causing the location of surface defects to shift and become elongated. Corus utilized the data for rapid quality monitoring (Descriptive) and launched improvement projects addressing the most common customer complaints (Predictive).

CQM developed the Infra-Monitor application in collaboration with ProRail. This application allows for the visualization of railway infrastructure and timetables in combination with other data sources (VARIETY). Using this detailed information, users can address various issues through interactive data analysis (Descriptive). One example is the Route Risk Register analysis. This analysis makes it possible to systematically identify infrastructure locations where the risk of a collision between two trains is above average. The analysis also provides insight into the impact of measures such as flank protection switches, ATBVV (Automatic Train Protection with Speed ​​Monitoring), and overrun distances (Predictive). By subsequently visualizing the results in a schematic track layout, experts gain immediate insight into the measures and their associated consequences. ProRail has already utilized the Route Risk Register analysis for various safety studies.