Big Data, Big Problems

Update : 13th May, 2024

SkyCell's CTO and co-founder, Nico Ros, is sharing insights on handling big data and using it to make better decisions that can reduce risk, cost and CO2.

These days, we have more data than ever. With over 328.77 million terabytes of data created each day, the potential for leveraging data to address industry challenges seems limitless.

While the goal of reducing costs and risks using this data remains, rising pressure now adds CO2 as a third dimension to calculating the Total Cost of Ownership (TCO). Reducing CO2 emissions is not only an environmental need, but CO2 will also be a very significant cost factor. A larger Pharma company emits approximately 10 million tons of CO2. To compensate for this with an average cost of $100 per ton, it will cost $1 billion per year. Therefore, reducing CO2 is a very effective way to save costs.

Cost + Risk + CO2 = The New Formula for TCO

Is it possible to reduce cost, risk and CO2 at the same time?
How can big data help and what are the problems we need to solve?

Firstly, it's hard to assess the quality of all data we need to determine its importance. What is Master data and what is dependent data? Inputting the wrong master data will result in the wrong output.

Secondly, the challenge lies in the modeling of data. Flawed methodologies, simulations, correlations, or assumptions lead to wrong decision-making and have knock-on effects on the entire supply chain.

Reducing cost, risk, and CO2 at the same time is challenging. Already only optimizing the cost is complex. For example, buying cheaper packaging that needs more volume in an aircraft can lead to higher airfreight costs and higher risk. At the same time, CO2 emissions increase because the new packaging has more volume and is further away from the optimal weight-volume ratio.

To reduce Risk, Cost, and CO2, this equation becomes three-dimensional and exponentially more complex.

Using Data to Make Better Decisions

So how do we find a solution that minimizes costs, risk, and CO2 together? At SkyCell, we use what we call the S+O approach.

This framework integrates simulation data – what could happen – with operational data – what actually happens. S+O data helps us make increasingly accurate predictions which maximize cost efficiency while reducing risk and CO2.

Operational Data

For operational data (O data), we need to look at what data points should be prioritized and how to collect high-quality data. Our research at SkyCell has led us to the conclusion that the key data point is position.

Positional data is the most important master data, a lot of other data depends on the correct position. With the right location, we can enrich our dataset with airport data, weather information, etc.. However, if our position is incorrect, all this data doesn’t apply. For example, if you believe the product is in Frankfurt but it is in Dubai, all data enrichment and connection with your shipment is worthless. Furthermore, if you rely on data obtained for weather updates, flight plans, and information about involved parties from Frankfurt, but your product is in Dubai, the data becomes irrelevant.

Meanwhile, if our positional data is correct but one enrichment data point is incorrect, our overall dataset is still good.

Collecting high-quality positional data manually can be complex and challenging, but with a network of IoT devices, this can become straightforward. At SkyCell, our containers and loggers enable us to have end-to-end visibility over the location of our shipments in real time, leading to accurately enriched operational data.

Simulation Data

For simulation data (S data), we need the right models. The key is collaboration. For every problem, someone out there knows the right model to solve it.

For instance, each pharmaceutical company generates vastly different numbers for CO2 emissions, with the results differing by up to a factor of 20. This inconsistency often stems from the use of overly simplistic calculation models, resulting in significant discrepancies from the actual emissions. The accuracy of these calculations is highly contingent on the model utilized.

The Centre for Transport and Logistics (CTL) at MIT has investigated this in detail to find the volume-weight ratio closest to the real output, which gives us higher-quality data models. Therefore, for data modelling and simulations, it’s a matter of finding the right team and partners to work with.

Once we have accurate simulation and operational data, this provides us with detailed insights into the supply chain so we can start balancing cost, risk, and CO2.

Combining S+O Data with AI

Although our S+O data approach allows us to make increasingly more accurate predictions over time, this can be difficult to do quickly at scale. Nico posited that we can now take one step further with AI. With the right high-quality data and models, AI can process hundreds of situations and find the optimal lanes much faster than a human can.

By leveraging AI algorithms together with S&O data, one can process vast datasets and answer what lane has the least risk, least cost, and least CO2 overall, turning big data into big opportunities.

SkyMind Decarbonize CO2 Calculator

As part of our commitment to collaboration, Nico unveiled our new CO2 calculation and reporting tool as part of our SkyMind DECABONIZE module. This free reporting tool will be released soon and offers a simple and impactful way to measure your logistics emissions and pinpoint areas of potential reduction. It uses the calculation method from the CTL at MIT and NTM (Network for Transport Measures), which is one of the best methods to calculate truck and aircraft emissions.

To discover more about this tool, sign up for SkyCell’s webinar and gain exclusive insight into its innovative features.

Source : SkyCell


This news article complies with our Editorial Policy

SkyCell News

Follow CCP