Winning the Losing Battle with Machine Learning and Artificial Intelligence
The best thing you can do is use existing technology to increase the chances of returning found items.
Without modern Software, an average found item costs your staff 45 minutes
With COVID-19, companies are looking for better ways to cut hours and save costs. Lost and Found is, unfortunately, a cost center for most. But it is important for your brand and reputation in so many ways (learn more about Customer Relations here: https://medium.com/lost-and-found-software/do-your-customers-love-you-4084c745f057).
Current Lost and Found Solutions were often made by IT departments or IT providers years ago. That’s why the solutions often reflect the state of technology and business practices as it was operated decades ago with virtually 0 innovation or further developments. The result? You can only add or search for items in what is simply a fancy online spreadsheet — there is no real strategy to save time or cost for both the company and the customers. The last “revolution” for many providers was to add an image of a found item within their system.
The goal of every company and its Lost & Found should be returning more items faster and easier for everyone involved. This approach leads to happy customers and saved costs.
The only way to achieve this is by utilizing “smart” technology — e.g. like our Matching Algorithm powered by Artificial Intelligence (AI) as well as our Image Recognition powered by Machine Learning (learn more here: https://medium.com/lost-and-found-software/lost-and-found-powered-by-image-recognition-602ba382d490)
AI is a game changer and a real timesaver
Let’s say the goal, for now, is to reduce the time it takes to register and identify the owner of lost property. Typically, we have two parties involved in this Lost and Found process.
- The party who found an item
- The party who lost an item
The Past
Whoever found the item registers and tries to describe it. With free text, dropdowns, and all the typos in the world.
The owner looking for their item repeatedly calls the (potential) finder or/and additionally writes plenty of emails with lots of free text.
The issue here is that a human needs to match these two unstructured types of information. That’s definitely not a time saver, nor is it efficient, fast, or practical. The result? More often than not, frustration for everyone involved.
The Present
To get everyone involved on a common level of understanding, we need to standardize the vast amount of information available. Once structured, we are now able to execute a highly dynamic and flexible system that makes it easy for a Smart Matching algorithm (like the one implemented in our Lost and Found Software) to identify owners for your staff.
To get AI involved, the system needs to identify the similarities between the remaining free text information that both parties entered and — a lot more complicated — identify any possible mistakes and typos made, i.e. iPad vs. eyepat
Let’s show an example:
Sam traveled from New York to Seattle. They forgot their laptop bag containing a tablet, chargers, notes, and a kindle at Seattle Airport.
With Lost and Found Software, Sam can easily submit (and later update) all items to their lost property report right from the lost and found page of the Seattle Airport website.
Airport staff registers the found items with state-of-the-art image recognition technology. Category, brand, color, etc. is automatically recognized and staff then adds any additional information. This process accelerates the registration process by up 90%!
Thanks to the standardized input on both ends, another real time saver is the matching algorithm. In seconds, it automatically checks and compares everything an employee had to manually in the past. Item type, name, date, serial number, place of loss, etc. The matching algorithm accounts for ALL items Sam submitted.
Imagine what happens when Sam goes to the airport’s website to update their case because they remember an additional item lost or more details about the loss. Staff would have to recheck the case after every minor edit, again and again, every day, for thousands of inquiries. So best let technology do these tedious tasks for you!
The Software matches lost property and found items automatically in real time.
In case you read “matching” in a Software Feature List, it’s worth taking a deeper look at what is actually meant. There are different possibilities: Maybe the matching only consists of checking the dates and categories, which can lead to hundreds of senseless matches that your staff then needs to confirm. This leads to the opposite effect of what was intended — MORE work for staff instead of less.
In contrast, the Lost and Found Software’s matching algorithm is changing the way of how we all work and return items dramatically. A “Matching Percentage” considering vast amounts of information is given to every item and inquiry. The algorithm even checks for patterns in free text, etc. That way, you get the best results possible.
The Future
Do you want to learn more about what’s coming next? For example, read all about “Real-Time Object Detection” here: https://medium.com/lost-and-found-software/lost-and-found-real-time-object-detection-820ebbdb078e
Are you interested in learning how to implement the leading, most innovative and advanced Lost and Found Software?
Read more here: https://medium.com/lost-and-found-software/how-long-does-it-take-to-implement-a-good-lost-and-found-service-c3c7e86350ee
or visit www.lostandfoundsoftware.com
