The statement “it’s a small world after all,” is no longer just a part of a song. It has now become reality, thanks to three ingenious inventions: the Internet, airplanes, and cars. Among the three, only planes and automobiles actually get you from a map’s point A to point B. But the more prevalent of the two modes of transport are cars.
According to a 2018 CarsGuide article, there are around 1.4 billion cars on the road worldwide. In the Philippines alone, there are about 1.1 million privately-owned automobiles. This means that millions to billions of vehicles could break down and cause safety issues if their owners did not know much about preventive maintenance.
The problem is worsened for second-hand vehicle owners, who may not have the operating manual of their cars because the original owners lost it. Fortunately, for this case, AutoServed has designed an app that helps car owners detect what parts of their vehicle need repair after a certain distance travelled. Our Data Science Fellows, through their capstone project, were able to help AutoServed improve the way the company ensures that its clients keep their vehicles in safe running condition.
AutoServed is a tech startup which has designed a revolutionary online app that enhances the way automotive preventive maintenance is carried out. Through this unique software, a car owner can do many things like find the most reasonably-priced estimates for different types of repair and pay for the relevant auto parts and labor.
But among the most wonderful features of the AutoServed app is the way it relates a vehicle’s mileage (the distance a vehicle has traveled) to the particular type of repair it needs. For instance, a certain distance of travel means the radiator has to be checked while another distance calls for an oil change.
Our Data Science Fellows Patrick Nuguid, King Calma, Justine Padayao and Andrei Gabriel Labayan were able to add value to AutoServed’s app through an innovative data science project. Their data-driven creation is a requirement for them to graduate from our part-time, 12-week-long Data Science Fellowship. Their work in improving the recommender engine for the startup’s online software helped them gain industry experience while indirectly fostering the growth of the auto repair industry and road safety.
Whenever a car breaks down, many mechanics and car owners suffer. Those working in auto repair shops sometimes can’t find a steady stream of customers. Some automobile owners, on the other hand, take a long time to find the right auto parts and mechanic for the right quality and price.
This is where AutoServed comes in. The startup’s app connects auto repair shops with owners who need estimates on repairs. The software also has a recommender engine that advises owners what parts need replacement after a certain distance that a car has traveled. However, the app can only make recommendations if a car owner has had a repair history via the AutoServed app or a car has traveled a certain distance (for example, older cars).
Our Data Science Fellows, however, had an idea that helped enhance the comprehensiveness of AutoServed’s app.
Using the volumes of car repair transaction information contained in the database of AutoServed, our team of Data Science Fellows was able to design a recommender app that can make repair recommendations even for cars that are either relatively new or those that have no service history through the AutoServed app.
Car owners input a lot of information on the app whenever they have their vehicle repaired through the software. Among the many types of data encoded by an owner include the distance his or her car has traveled before entering the repair shop (what is referred to as “mileage”), the car’s brand and model, the year the automobile was manufactured, and the broken parts that need fixing. AutoServed stores thousands of these repair transactions and their relevant car-related information, ready for processing.
The first thing that the team of Data Science Fellows did was to process the user data to make them clean and balanced. This process is called oversampling. For example, if a transaction is missing information like the year the car was manufactured, the Fellows searched for the car’s Vehicle Identification Number (VIN) and from this information estimated the year the car was manufactured. Other ways of ensuring that information is balanced include correcting spelling errors of car brands or models and/or deleting any transaction that has missing information.
Once all this car repair information had been processed, our Fellows used the Random Forest Classifier algorithm to process the volumes of data they cleaned. This algorithm, which powers the team’s improved recommender app, predicts what an owner’s car should have had repaired given a certain past mileage.
To illustrate how the team’s recommender app works, let us have a look at the example illustrated above. Here, an owner has a 2016 model Ford Everest. According to his vehicle’s odometer (a device that measures the total distance a car has traveled), his vehicle has covered 188,124 kilometers. All the owner has to do is input this number, along with his vehicle’s brand, model and year manufactured. The old AutoServed app has the capability to recommend future repairs like those involving car parts like the air filter and the cooling system (the two components must be repaired when the vehicle’s odometer reaches the 206,936 kilometer mark and 216,342 kilometer mark respectively).
However, thanks to our Fellows’ improved recommender app, the startup can now make repair recommendations for past mileage figures. For instance, in this example, the owner should have had his tires and wheels checked or replaced when his vehicle reached a distance of 150,500 kilometers. In other words, the owner has to have a look at his tires and wheels because his vehicle’s current mileage (188,124 kilometers) is way beyond the 150,500 mark.
Our Fellows have applied data science in the field of automotive maintenance. Though this in itself is already an impressive achievement, some of you may be wondering, “Where else in the automotive industry does data science impact?”
When the Ford Motor Company launched the Model T in 1908, it was hailed as an innovative product which changed mobility access for many. Today, the automotive industry is still at the cutting edge of innovation. But rather than innovating through manufacturing, innovative features are driven by data science.
Auto repair is just a link in the chain of the large economic sector called the automotive industry. The most commonly mentioned data science application in the automotive industry is perhaps the rise of connected and autonomous vehicles, which combines deep learning models and Internet of Things (IoT) sensors.For instance, self-driving cars are expected to generate and consume around 40 terabytes of data for every eight hours of driving and this data is going to be significantly more than the amount of data generated by an average person today. The sheer volume of data to be generated by the automotive industry gives value to the use of data science across many more use cases.
However, some of you may ask: What if I don’t like cars or don’t want to work in the auto industry? What can I gain from learning data science?
The automotive industry is not the only sector of the economy which is affected by data science. The Philippine economic planning agency NEDA says in a 2021 Philippine News Agency article that “data science...can help make better policies and deliver better services as the Philippines heads to a new and better normal.” In other words, even the Philippine government is pushing for the development of data science.
Therefore, whether you go for the auto industry and other types of businesses, or decide to serve the government, skills in data science can help propel you forward towards career success. After all, the global demand for data scientists is still growing despite the pandemic. According to a 2021 CNBC article, this increase in demand is partly due to the increasing popularity of work-from-home setups.
If you’re dreaming of launching your career into high gear, or want to develop your organization or company, you may consider enrolling in our Data Science Fellowship. This upskilling program is unlike many others in the country because it is engaging, industry-relevant, and project-based. Like our Data Science Fellows who helped AutoServed, you will also create data-powered projects with the help of an industry mentor. Our mentors work together with our instructors to ensure that no learner gets left behind.
The perks of our data upskilling programs don’t end there.
We are happy to announce that we have just launched an additional component of our data learning tracks. Called the Industry Apprenticeship, this program gives our students the chance to get paid while continuing the projects they started in the bootcamps they enrolled in. Thanks to Accenture and the startups of the Asian Institute of Management - Dado Banatao Incubator (AIM-DBI), our next cohort of Data Science Fellowship learners will get the chance to earn while working on real-world business problems. If you want, you can learn more about the apprenticeship by having a look at the links in the Recommended Reading section below.
If you are a company, institution or organization looking to bring data talent to your team or build up the skills of your existing workforce, you can also reach out to us at email@example.com.
Through their recommender app, our Data Science Fellows helped AutoServed indirectly contribute to road safety through effective vehicle maintenance. In a way, the startup company has also contributed to the economic well-being of both mechanics and car owners. This capstone project is a testament to how we at Eskwelabs help promote data education for the common good. This mission powers us to provide innovative data education for as many people as possible.
If you partner with us, we can be the engine that can help power your career or organizational growth to prepare you for the future of work.
Interested in our Industry Apprenticeship? You may click any of the links below to learn more, according to your program of choice.
Let us teach you how to create your own recommender app and other socially-impactful data projects. Click here to know more about or sign up for our Data Science Fellowship.
If you are more interested in creating interpretations and visualizations of large chunks of data, then our Data Analytics Bootcamp may be more your speed. Proceed to this page to learn more or sign up.
Data science helps agritech startups improve the lives of farmers and consumers. How? You may read this blog to learn more.
Want to know how our Data Science Fellows helped make legal research more accessible and high-tech? Read this blog post to find out.