Research for a data science article synopsis or blog post that details a complicated real-world problem that was solved (or remains unsolved) using data analytics or predictive analytics.
Include details about the problem space itself, the technologies that were used to solve the problem, and what difficulties they encountered. If the problem remains unsolved, be sure to mention why and the obstacles that haven’t been overcome.
Provide a link to the article/blog post you are discussing. Finally:
Source: https://www.business-standard.com/article/companies/2022-to-see-rise-in-adoption-of-chatbots-hiring-in-data-analytics-report-122011001150_1.html (Links to an external site.)
One problem solved by data analytics is chatbots and remote work opportunities. From this article, we can see that chatbots are the solution to customer service needs. We probably all encounter chatbots nearly every day. The utilization of Chatbots allows for a better, more focused customer service experience. From my experience, there are chatbots that can pick up on keywords and phrases and yield an answer that is fitting to what the consumer is searching for.
UPS created a handheld route optimization tool called ORION (Links to an external site.). ORION provided drivers with the most efficient route for deliveries and pickups. To create this route UPS had to make custom detailed maps. The reason they couldn’t use map services like Bing or Google is because some drop off and delivery locations are not addresses but they are locations. Over time they have enhanced ORION to provide real time updates based on traffic and unknown pickup/drop off commitments when the original route was planned.
After collecting all of this logistical data, UPS was have to review the data and provider even further optimization. In 1959 George Dantzig created the vehicle routing problem which can be used to organize many things. For UPS finding the shortest route was instrumental but later finding the most optimized route which shortened total drive time became the most important. What UPS found after reviewing the data is that left turns have a much higher risk of accidents and a much longer wait time for traffic to clear to make the left turn. UPS found that if they optimized their route using right turns and only taking left turns when absolutely needed, it saved 10 million gallons of fuel, avoided the emissions of 20,000 tons of CO2, and delivered 350,000 more packages a year. The optimization model allowed for a max of 10% left turns in the route. As a result in this new optimization UPS was able to reduce the number of trucks by 1,100 and total distance travelled by 28.5 million miles. For a company it’s always great to save money and have a better customer experience.
This is a beautifully described use case from UPS on how we can achieve the most out of Data analytics and find an optimum route for delivery. With the power of analytics, UPS has dramatically changed its mode of operation.
Here is the takeaway from the blog:
Hierarchy of Impact through the use of analytics
Descriptive -> Diagnostics -> Predictive -> Prescriptive
Note: With the growth towards the more advanced feature of analytics the business also grows.
UPS was overwhelmed with great results with the help of analytics. They were able to save 100M miles driven, 8M gallons of fuel saving, 95 % reduction in load training time, 8B fewer manual entry, and 100,000 metric tons of carbon emissions
We should be able to forecast and plan ahead of what would be our demands and how we can meet their demands. With the use of Prescriptive analytics and by the use of the analytical tool and advanced algorithms we can take an even better decision (rather than taking risks in taking a decision )
For e.g.: ORION helps the UPS drivers to assist them on which route to go and which order to deliver in order. All the driver needs to do is follow the algorithm. It helps in route optimization.
And with the help of prescriptive analytics UPS was able to save an additional 300 to 400 M dollars annually
There are still many open questions that UPS is trying to figure out by use of data that has already been collected and analyzed it like if they need more access points, lockers, etc.