The road to mobility market intelligence

How to generate value through predictive forecasting

An ecosystem full of data

The mobility market – just like any manufacturing industry – must constantly walk the thin line between market supply and demand. Investing on incentives in order to reduce inventory is not always the right solution when the supply cannot meet the demand. A lack of vehicle production also increases the risk of leaving clients dissatisfied or – even worse – losing their brand loyalty.

Together with this struggle, OEM and dealers are left juggling new consumer trends. Customers are often intrigued by the diversified options that new technological developments are bringing to the market. But there is more. Stricter governmental regulations with a focus on sustainability have created a shift toward a mobility-as-a-service model in which autonomous vehicles are just one of the many and new alternatives developed to adhere to laws and legislations. The increase of fleet market sales has also resulted in an increased number of concepts – as well as new products – for markets such as shared mobility or connected cars.

This multi-options market has not only paved the road to endless technological possibilities but also opened the doors to an ecosystem full of data that can – and must – be explored. But how can we best juggle with the enormous amount of information produced every second in the mobility industry? The answer lies partially in the concept of predictive forecasting itself. This is one of the best tools for cleaning up this mess of data and shaping it into something of real value that can have a positive impact on the automotive market. Let’s look at this concept in more detail.


How does predictive forecasting work?

In layman terms, the technique involves the use of statistical models (which most frequently include predictive models) to analyze historical data and make inferences based on the results. One of the central goals of the methodology is understanding whether business-related goals are aligned with the inferences/predictions resulting from these analyses. Thanks to the resulting conclusions, reliable orientation guidelines for actions can help companies to stay on the right track and/or learn from their mistakes.

Would you like to look at the whole idea from a less technical point of view? This is nothing more than a methodology for predicting the future. Quantitative analysis is used to find the answers to questions such as “How long will a mechanic component survive before needing replacement?”, “How likely is this customer to purchase a new vehicle in the next three years?”, or even “What are the hot spots in the area for a taxi driver?”.

Let’s look at what this step-by step process of predictive forecasting will look like in the mobility market.

The workflow of a predictive lifecycle

To a certain extent, predictive analysis is fairly similar to a data lifecycle. Instead of information simply being archived or destroyed, it is transformed into something else after it is retrieved and processed. The processes and models that manipulate the raw data create a product from this information. This, in turn, generates added value and – together with it – a new way of looking at the knowledge.

In specific, these mechanisms consist of:

  • Retrieving data from relevant data sources to import into the model. With such a multitude of information, the first step towards achieving a reliable forecasting process is understanding and selecting the right ‘insights’ from the right sources to include in the analysis.
  • Pre-processing the collected data to respect quality standards. Data input from data sources always starts as raw data. In order to ensure that it is reliable , it must undergo several operations so that it can be transformed into “ready-to-use” or, in other words, qualitatively acceptable information.
  • Developing a predictive model to process and transform the data. The pre-prepared data is ready to be interpreted. At this stage, the only missing piece of the puzzle is a set of rules to serve as guidelines that are trained to “ask the right questions” of the collected information in order to retrieve targeted answers: the predictive model.
  • Integrating the forecasting model into a production environment. Once a correct and suitable predictive model is found, it can go into production. Having the model in production means that it is now possible to provide the “services” to devices or software that require the aforementioned answers.
211216_Predictive Forecasting-EN

What does this mean for the mobility ecosystem

The introduction of predictive forecasting in the automotive market has opened the doors to new possibilities that have changed how we look at the industry and transformed it into an intelligent digital mobility environment.

  • Increased support for the company’s planning and performance. The prediction and analysis of supply and demand through forecasting models eases the stakeholders’ market evaluation of a company as well as the adaptation to the changing market environment.
  • More reliable factors to count in the decision-making process. Automotive firms can now sharpen their decision-making abilities through in-depth predictions on targeted requirements and/or problems.
  • 360-degree view of operational ecosystem. The analysis of sensor analysis not only provides an overview of the vehicle but can also paint a picture of how the element reacts to or is affected by its environment. A similar mechanism happens in the value chain of the mobility market in which the single players (e.g., customers, OEMs, and dealerships) can be analyzed either as individuals or in relation to the industry ecosystem they are interacting with.
  • Understanding and alignment with the needs of stakeholders. The different parties often have preferences, needs, or requirements that aren’t entirely dependent on the market environment and must therefore be treated individually. Forecasting models can offer support in identifying these factors and provide a beneficial stakeholder-specific approach in order to take on stakeholder-specific problems with stakeholder-specific solutions.

These advantages are just an introduction to the first steps that not only the mobility market but also many other industries can take in working towards new value creation in their business ecosystems.

Let’s now  look at how these vantage points enter the ecosystem and shape the mobility industry through some hands-on use cases.

Where is predictive forecasting in the mobility industry?

  • Inside a dealership sales software. Where data from customers, purchased vehicles, and configuration preferences is input. The predictive algorithm helps sales representatives detect changes in consumer trends and identify target customers for a new vehicle purchase. The leverage of data power can reach the point of learning that one particular customer is likely to purchase a black sedan with winter tires within one year. Knowing this means knowing which direction to follow in order to either convert a non-customer to a buyer or provide an existing one with a newer, more coveted model.
  • On a truck, in an autonomous vehicle. Particularly (but not only) in the case of logistics vehicles, which often transport valuable cargo, harsh weather conditions such as rain, storms, and snow can significantly influence the perception and navigation performance. On top of this, the behavior of other drivers can often be unpredictable. A new sign or an accident can drastically change how vehicles move on the road. In order to partially mitigate these conditions, predictive models support the driver by detecting as much of this information as possible in advance  in order to reduce the collision hazard and optimize the driving performance.
  • In a mobile app. Where taxi drivers can learn where the “hot spots” for pick-ups are. The prediction based on data collected helps drivers distribute themselves evenly and optimize their fares. They now not only avoid long waits between clients but can also select the “hot spots” that are closer to them, thereby leading to fuel savings.
  • Inside an automotive equipment manufacturer factory. Predictive analytics can help estimate the expected supply and demand of parts, improve manufacturing quality, and anticipate the needs of the production cycle. A machine operator can now have a clear idea of the life expectancy of a machine and can thus properly carry out maintenance or plan for a replacement. In this environment, employees and machines can now work transparently in an ecosystem in which forecasting is the booster for efficiency in operations and planning.
Cloudflight implemented a forecasting system based on existing data from the data market Austria to predict demand for taxis at interesting locations in cities.

Keeping up with market speed

The new technologies and regulations as well as approaches to the world of mobility is transforming its meaning at a speed that is often hard for markets and companies to catch up with. The market changes – together with the trillions of data points produced every day by vehicles, sensors, and devices – require ad hoc solutions so that companies can continue doing business and keep up with  the changes in the ecosystem. Leveraging the power of predictive forecasting leads to increased control throughout the entire supply chain and provides a tool for car manufacturers and dealerships to be able to use insights from data in order to predict the needs of customers.

Anticipating events – whether the need for a car, the maintenance of machinery, or the risk of floods – is the best way to prepare for them. And it is precisely what predictive analytics provides. Having a clearer idea of the “what ifs”, forecasting indeed provides a new tool for reducing risks and leveraging an earlier unknown factor derived from the complicated world of data.

Would you like to know more about predictive forecasting? Get in touch with us. We’ll be more than happy to show you how Cloudflight leverages this technology to support its clients in the mobility market.


Would you like to know more about predictive forecasting?

Get in touch

Get in touch