Kardinal: Optimizing last mile delivery by unlocking value upstream

Jonathan Bouaziz, CEO and co-founder of the last mile delivery management platform, Kardinal, discusses the importance of working on cargo and delivery area adaptability in a time of major business fluctuations and declining margins for carriers.

How does the economic situation impact the activity of your clients, last mile operators?

The parcel delivery industry is impacted by tremendous fluctuations in parcels quantities. After a sharp increase in the last few years, driven by the boom in e-commerce and the Covid-related lockdowns, delivery volumes are now down at the end of the year because of inflation. On top of this, there is also a shortage of drivers and supplier failings, whose costs have risen sharply in recent months, especially with regard to fuel. Energy and warehouse maintenance costs have also increased. The demand for more environmentally friendly logistics from end customers also has a cost, as does the improved quality of service. This situation leads to a decrease in profitability.

How does Kardinal tackle these problems?

Any organization that is experiencing rapid growth and needs to optimize its costs must be able to make decisions quickly. To improve last mile delivery operations, carriers are trying to work on route scheduling from the moment packages are on the truck. However, we bring value upstream of this task. We help optimize the delivery area, from warehouse location to area sectorization, to know which goods go in which cargo load. We provide a strategic or tactical vision that helps optimize last mile delivery by unlocking value upstream because that’s where there’s the most potential.

Jonathan Bouaziz, CEO

We help optimize the delivery area, from warehouse location to area sectorization, to know which goods go in which cargo load.

Jonathan Bouaziz, CEO and co-founder of Kardinal

What type of data do you use?

We send historical data to the software, which analyzes it and divides the territory into small cells based on a given level of activity. We include delivery dates, delivery locations, whether operations are BtoB or BtoC, duration, size, volume, weight, available resources, resource type, different cost models…

As a result, we can generate heat maps that capture the level of activity in each of these areas, the total cost of the route and its resilience to activity changes. The algorithm we developed allows us to identify the links between these different zones and combine them to produce stable and reliable business operations. These cells act as areas from which we build a range of future scenarios based on five use cases: territory optimization, fleet conversion, resource planning, warehouse location, and route optimization.

Users can use these recommendations to test new operational strategies and become more agile in their decision making. For example, in France we rolled out the solution to all 72 DPD depots. All depot managers are now using Kardinal.

How often is the data updated?

It depends on the IT experience of our clients and the frequency at which they upload the data. Some do it once a day, some once a week, some once a month. But the efficiency of the areas defined by our tool is not based on this frequency. The business activity of the past year, on which we usually build our projections, does not exactly determine what the future will bring. The secret of our algorithm is to consider this history as a relative truth from which it will build an infinite number of scenarios. Hence, when a given situation occurs, it will automatically match one of these scenarios built by the solution.

Is data quality also something that can weaken your model?

This is a very important question, indeed it can be complicated to build a reliable model if the data is not sound enough. Therefore, we developed new techniques that allow us to work with less-than-perfect data. With 85% reliable data, we can build a solid operation model over time. We are aware that the picture we have of past data is not entirely true to reality. Our algorithm therefore works to build a picture that is as close to reality as possible.

Our approach consists in forecasting business operations by generating an infinite number of business scenarios.

How far has the company grown after raising funds in 2019 and this year?

We’ve doubled the size of our teams in just one year. There are now forty of us working at Kardinal, including five PhDs in mathematics. We’ve launched pilot programs to expand in Europe, starting in Germany, England, Belgium and Spain.

This is an interview of Jonathan Bouaziz made by Républik Supply, a French Supply Chain media.