Digital Twin Genie Case Study: 54% reduction in automotive manufacturing costs

Key points

    Specific client goals and objectives

    Explaining the steps that were executed

    Examining the final results and outcomes

By Carlos Miskinis
Digital twin research expert
Mar 2018

Privacy Notice:

The case study that will be examined below will exclude specific information that is labeled as private. Some percentage of our case study content has been purposefully restricted in order to protect the privacy of our clients. Public access will not be granted to sensitive information such as company and/or employee names, specific business metrics and internal process descriptions in order to fully comply with the rights and demands of our clients and/or partners that chose to work with Challenge Advisory.


Project overview and KPI establishment:


Digital Twin Genie has been utilized with great success as an effective tool for minimizing manufacturing and development costs of our client in the automobile manufacturing industry. The specific goals that were put into place by the client itself were to improve the marginal profits (the percentage is listed below) per car launch in order to achieve a new floor/base of stable profits. Moreover, although moderately optimized, the overall speed and agility of their car manufacturing process were improved to an average of 14-17 hours throughout a 15.5 month period that was managed and arranged by their internal department of manufacturing.


The objectives included the task of improving the average time it takes to manufacture a car from 14-17h period down to 12-13 hours, as this would allow them to manufacture 238 extra cars every single quarter, increasing their profits by an average of 15-20%. Finally, in order to achieve better product to end-consumer interaction, they required a strict set of new metrics to be achieved to gain better and more transparent insights into how their customers are utilizing their product. According to their perspective, Digital Twin Genie was a suitable and cost-effective solution for achieving these metrics and according to our stipulations, the software managed to meet and exceed their expectations – the results and outcomes will be elaborated at the end of the case study.


The final objectives and targets were set as follows:


  • Increase overall car manufacturing profit margins by 15-20%
  • Reduce the average manufacturing timeframe down to 12-13h
  • Gain further insights into how their customers interact with the finalized product (e.g. weekly mileage metrics, average fuel consumption, etc.)


When it comes to key performance metrics such as profit margins, the biggest hurdle that needed to be overcome by the client was the ever-increasing development and manufacturing costs that were a huge hindrance for growing profit. On the other hand, it was their most important, key internal process that was paramount for launching new model designs and testing their product performance. This was the biggest challenge from the very start of their newest model’s launch as the company was forced to meet their set deadlines from the launch year going forward.



Moreover, considering the pressing challenges that the automobile industry is facing for the long-term, such as adopting new solutions to improve sustainability and adopting a more eco-friendly approach in terms of automobile usage that caused a spike in electric car development, the industry’s competitiveness is forming a substantial gap between fuel-powered and electricity-powered car manufacturers. This causes a direct need for automobile manufacturers to stay on the cutting edge and ensure their brand’s uniqueness, by engineering new solutions such as Digital Twin to enhance productivity.


One of the biggest advantages of fuel-powered car manufacturers is their popularity due to the innate preference the average car driver has, which usually tends to gravitate from electric car manufacturers altogether. However, the competitive landscape is getting harsher and the internal board of our client’s company has made a firm decision to adopt Digital Twin Genie as their first solution to test – we will cover how Challenge Advisory has implemented the changes necessary to help the client reach the objectives they’ve set.

Part 2

Defining our step by step approach for implementing Digital Twin Genie

The most important factor that makes digital twin genie work is data acquisition - the most essential part of the entire process

Digital Twin Genie has 5 foundational steps that enable data collection and aggregation to happen. Data is a necessary part of the process due to the fact, that every single move and decision will be based on it. The very same moment a digital twin of a product or an intangible service is formed is the moment where the overall success and accuracy of a Digital Twinning campaign will be determined. Apart from being the world’s first cost-efficient solution that is applicable for any industry out there, Digital Twin Genie has a few unique features that allow it to become more superior than other Digital Twin data modeling or software solutions in the market.


Those key distinctive features of Digital Twin Genie include:


  • AI-powered machine learning (gives developers the ability to foresee future failure conditions and avoid them)
  • Remove data aggregation (allows for data stacking to continue even after the virtual twin has been disconnected from its physical counterpart)
  • Database-centric recommendations (takes over decision-making and bases it on a fully automated basis after enough data has been collected)


During the total timeframe of the project (14 days) that was dedicated to meeting the initial objectives of our client in the automobile manufacturing industry, we have initiated the installation of Digital Twin Genie. Excluding the multiple meetings that were solely dedicated to covering our A-Z client-centered approach and strategy, the very first core action that was taken was the acquisition of the development kit, which is presented directly to our clients at the very beginning of each project.



However, in this case, the solution had to be completely unique and bespoke in order to fully meet the needs of the company we were working with. Meaning, that the Digital Twin sensors themselves had to be altered in order to have the capabilities of not just tracking the performance and manufacturing speed of the products themselves, but they also had to be specifically built into the core components of their car engines, in order to successfully monitor their performance and gain full clarity on the full spectrum of interactions the customers engaged in with the products themselves.


The second step was the installation of the Digital Twin Genie software into their internal team’s computing devices, that were responsible for controlling the main processes of manufacturing. This helped our Digital Twin experts gain the ability to monitor their equipment and gather data on vital processes, including:


  • Machinery temperature
  • Performance speed
  • Power input and output


The specific sensors that were used to gain access to these metrics consisted of analog and digital sensors, including temperature, IR, ultrasonic and proximity sensors – all built as bespoke equipment made specifically for the client itself. Moreover, these 4 main types of sensors had to directly cooperate with the software itself in order to make data aggregation available.


In order to display the monitoring capabilities of the software more visually, below you can see an example that was directly taken from the Digital Twin Genie’s dashboard itself. The entire dashboard of the software was not captured due to the privacy policy that was established by the client that you can read at the very beginning of the article. However, regardless of the imposed restrictions, the photo below does a valid job at giving you a good visual on how some parts of the interface look like:



The third step revolved around successfully starting to accumulate data from the manufacturing equipment and determine the speed and motion angles using which the company’s tools performed. The motion and proximity sensors were heavily utilized in order to achieve the information we needed and we are going to explain how acquiring all of this information helped us to greatly exceed the expectations of our client.

How Digital Twin Genie exceeded the initial set objectives of our client

Discussing the momentum that was gained from data acquisition and aggregation

The overall timeframe of the project that we required in order to reach the objectives for our client took 2 weeks. As we’ve discussed in the first two parts of this case study, the first week was dedicated to the setup, installation and data acquisition that the Digital Twin Genie platform needed in order to make the campaign. With all the preparation and custom sensor alterations in mind, the first part was completed in 6 days, however, the official timeframe for part 1 was announced as 1 week. The second week was dedicated towards additional data aggregation and stacking, which allowed our software to calculate how the speed of the machines and the profit margins could be increased. Now, since the final stage of the entire process is covered, let’s get into the results:


  • Automobile manufacturing profit margins were increased by 41-54% per model
  • The estimated average automobile manufacturing time was reduced down to 9-10h
  • The depth of customer interaction was expanded with additional features such as gas consumption, daily mileage, and engine performance monitoring


Although seemingly impressive, the results that were displayed above didn’t represent the best case study that Challenge Advisory has managed to acquire by using Digital Twin Genie. The capabilities the software has due to its machine learning engine that is powered by concepts partially designed by Google (our strategic partner) exceed all current Digital Twin solutions that are in the current market. In addition, we will continue to improve the software, putting our main focus on the accuracy of the collected data.



Quickly summarizing and explaining the results and outcomes of Digital Twin Genie


The total average profit margin increase of 41-54 percent:


The biggest factor that gave us the circumstances to implement such a huge spike in profit margins was the accuracy of data and machine learning capabilities of our software. Meaning, that since our solution was capable of tracking every single move of the equipment that manufactures car parts, we were able to gain full clarity on what moves and actions can be shortened and made quicker – this resulted in lesser electricity consumption and faster performance.


Secondly, the overall downtime of the machines was reduced by 37%. This happened because, in most car manufacturing plants, developers do not have the real-time, direct digital vision of their equipment’s performance compared to the accuracy that Digital Twin technology has to offer. This presupposes that if the machinery breaks, it will need to be repaired, preventing it from producing automobile parts during that timeframe. Digital Twin Genie allowed us to completely eliminate this failure condition by monitoring which part of a specific piece of machinery was about to experience failure, making it easy to be replaced during non-production hours.


Average production time reduced to 9-10 hours:


Meeting this objective was fairly simple in comparison to the other two goals we have been challenged with. The Digital Twin Genie software has a built-in beta version of its custom-made machine learning algorithm. It is important to keep in mind, however, that the premise of the ML program is different than AI. By gaining data about the performance of our client’s processes and tools, we’ve quickly managed to realize that there are severe gaps between specific points of interaction.


For instance:


The time it took for automated machines to travel from point A to point B took about 1.2351.267 seconds. By utilizing Digital Twin Genie, we have shortened the amount of space the tools needed to cover in order to accomplish the same outcome, down to an average of 0.70 seconds. From a perspective that is extremely zoomed in, this might not seem like a lot, however, the machines constantly perform thousands of motions throughout the timeframe of an entire workday. Because of this, the small speed difference is multiplied thousands of times – the calculations added up to exactly 9-10h that their new process required to manufacture a fully functioning car model.


Enhancing the number of insights received from active customers:


One of the newest features of Digital Twin Genie we have recently built in (the date of the last update was: 03/Dec/2018) is called: predictive data stipulation. The main use of this feature was to track the activity of the physical twin in the digital twin interface even after both twins were disconnected.


This update is the single most important reason that allowed us to give our client the extra insights they needed to figure out how their customers interact with their product, gain full transparency into the average weekly drive times that were incorporated in hundreds of car models and most importantly, what features their manufactured cars are lacking. All of these extra features provided the company with in-depth metrics that were calculated in real-time via their Digital Twin Genie dashboard.

More information