AI for Smart Parcel Lockers. Optimizing Routing, Scheduling, and Predicting Pickup Times in Last-Mile Delivery Process

AI for Smart Parcel Lockers. Optimizing Routing, Scheduling, and Predicting Pickup Times in Last-Mile Delivery Process

Dorota Owczarek - August 23, 2022

As we observe the exponential growth of online shopping, the last mile of delivery is becoming increasingly important. This is the final leg of the journey from the warehouse or store to your doorstep, and it can be a costly and time-consuming process. To ensure that packages are delivered on time and without issue, optimizing the routing, scheduling, and prediction of pickup times for smart parcel lockers is essential. With advances in artificial intelligence (AI), this process can be streamlined and improved significantly.

This article explores how AI can be used to optimize last-mile deliveries to smart parcel lockers.

Home Delivery vs. Smart Parcel Lockers

There are two main types of last-mile delivery: home delivery and smart parcel locker delivery. Home delivery is when a parcel arrives directly on your doorstep, while smart parcel locker delivery is when packages are delivered to a secure locker that can be accessed with a code or key. Both methods have their own advantages and disadvantages.

Home delivery is typically more expensive than smart parcel locker delivery, as it requires additional labor to transport the package from the truck to your doorstep. In addition, home deliveries are often subject to weather conditions and traffic, which can delay or even prevent the package from being delivered on time. Furthermore, there is a fair chance that they won’t be present at the delivery time, resulting in a failed delivery. Failed deliveries are costly for all parties involved, inconvenient for customers, and cause more strain on the environment than necessary. However, home deliveries are generally more convenient for customers, as they do not need to travel to a specific location to pick up their shipments.

On the other hand, smart parcel locker delivery is often more cost-effective than home delivery, as it requires less labor to transport multiple packages from the truck to the destination locker. In addition, smart parcel lockers are typically located in high-traffic areas, making it easier and faster for customers to pick up their shipments whenever possible. Smart lockers not only enable 24/7 pick-up options but also support drop-off and returns of parcels. Thanks to that, some part of the labor connected to last-mile delivery is transferred over to the customer.

The Increasing Popularity of Smart Parcel Lockers in Last Mile Logistics

While home delivery is still the most popular method of last-mile delivery, smart parcel locker delivery is becoming increasingly popular. This is because smart parcel lockers are more cost-effective, convenient, and eco-friendly than home delivery. In addition, smart parcel locker delivery is often faster and more reliable than home delivery.

Global market size of smart parcel lockers, current and projections for 2028

More customers use parcel lockers with the smart parcel locker market share predicted to double by 2028, reaching $1630.2 millions globally. Source: statistia

Artificial Intelligence for Smart Parcel Lockers

As the popularity of smart parcel locker delivery increases, it is needed to manage them more efficiently. AI can optimize the route, schedule, and prediction of pickup times for smart parcel lockers. By doing so, we can make sure packages are delivered on time and without issue.

Custom machine learning solutions for improvement of last mile delivery processes

Development process for custom machine learning solutions for improvement of last mile delivery processes

Dynamic Parcel Routing System

A dynamic parcel routing system is a type of software that uses artificial intelligence to optimize the route of a delivery driver. This system considers traffic data, weather conditions, and the number of packages to be delivered. By doing so, it can calculate the best possible route for the driver to take. This not only saves time but also reduces fuel consumption and emissions by limiting unnecessary driving.

There are a few different types of base route optimization models: the traveling salesman problem, the vehicle routing problem, and the pick-up and delivery problem. The traveling salesman problem is a classic optimization problem that involves finding the shortest possible route that visits each destination exactly once. The vehicle routing problem is similar to the traveling salesman problem, but it also considers the vehicle’s capacity. The pick-up and delivery problem is an extension of the vehicle routing problem that includes picking up and delivering packages along the route.

Distributing and picking up either undelivered or shipped parcels (retrieving parcels) back to distribution centers for the smart parcel lockers network is a critical but challenging task for the smart locker industry. Finding an optimal route for shipments to be delivered and picked up, considering vehicle capacity and traffic congestion, is a classic problem where machine learning brings a competitive edge.

Smart Delivery Scheduling

The problem of last-mile delivery has one additional dimension that adds complexity to the vehicle routing problem - time. In last-mile delivery, packages must be scheduled for delivery to minimize the number of courier visits to an individual parcel locker, optimize delivery times and reduce labor and fuel consumption needed to fulfill deliveries. This is known as the smart scheduling problem.

Smart scheduling uses artificial intelligence to schedule delivery times and suggest a delivery route for smart parcel lockers. This system considers the number of packages to be delivered, the location of the smart parcel locker, and the availability of delivery drivers. By doing so, it can determine when each package should be delivered. This not only saves time but also reduces fuel costs and emissions along the supply chain.

Predicting Pickup Times

Understanding how each customer interacts with parcel lockers is important to provide an optimal locker experience and minimize the cost of smart locker operations. The history of deliveries connected with a particular phone number/email account is a gold mine of information that can be used to understand customer behavior better and optimize the smart locker network.

What can the data tell you? You have pickup patterns connected to the parcel locker addresses. You have shipper/retailer data, end customer address, customer working patterns, or even understand whether they use a car or pickup parcels on foot.

Segmentation of parcel lockers is also crucial. You can connect certain addresses with domestic areas, shopping malls, communication centers, or office districts by getting a better understanding of where specific parcel lockers are located and how people usually pick up their deliveries. You can also identify popular parcel lockers or when more parcels will be sent to each locker. This infers the type of customer that is likely to use a particular smart locker.

Predictive analytics additionally take into account the time of day, day of the week, and holiday schedules. This data can be used to predict when a smart parcel locker will be used and how long it will take for a package to be picked up. This information can route parcels, schedule courier visits, and optimize the delivery process even further.

Related case study: Implementing AI model to optimize routes and timelines of deliveries

A company from the logistics sector approached us to create a custom AI model that optimizes routes and the scheduling of deliveries.

Our challenge? The key challenge here was to prepare a dedicated AI-based system designed for carriers to optimize delivery time depending on the destination address. Thanks to the model we managed to reduce failed and late delivery rates by 30%. Read our detailed case study of this project.

Predicting Locker Filling Rate (Parcel Locker Capacity)

Locker filling rate is another critical metric that smart locker delivery companies need to optimize. If a smart parcel locker is not being used often, it may be unnecessary to keep it in operation. On the other hand, if a smart parcel locker is being used too often, it may become full and unable to accept new shipments. Setting up additional modules or installing new parcel lockers nearby may not be the optimal answer.

Machine learning solutions can be used to predict locker capacity rates together with accurate forecasts of pickup times and adjust the scheduling of deliveries accordingly. Understanding parcel sizes and weights and connecting them with final lockers are also crucial pieces of information when planning parcel locker capacity. By applying historical data on locker usage, last-mile delivery companies can get more accurate forecasts, avoid filled-up lockers, and optimize their smart locker networks.

The information about the predicted locker filling rate can be used at the moment of ordering delivery. Customers could be asked to select a different parcel locker upon finishing their order if the prediction model forecasts limited space at certain parcel lockers. This kind of feature increases customer satisfaction by providing transparent information and proactively suggesting alternative best solution.

Predicting Failed Deliveries

Another important metric that smart locker delivery companies need to optimize is the number of failed deliveries. A failed delivery occurs whenever a customer fails to pick up the parcel within the limited time frame. No one is happy with a failed delivery. The costs of returning the parcel back to the retailer are high, and the customer is left without their purchase. Parcel locker capacity has limited space that has to serve more parcels, so you cannot extend waiting times indefinitely.

This is where artificial intelligence can help. By analyzing past data, smart locker delivery companies can predict which addresses are more likely to result in a failed delivery. This information can be used to schedule deliveries to these locations at times when it is more likely that someone will be able to pick up the parcel. In this way, smart locker delivery companies can minimize the number of failed deliveries and improve customer satisfaction.

Parcel Consolidation for Parcel Lockers

With the growth of eCommerce and the growing popularity of free-of-charge delivery services, we notice that some customers order multiple shipments within a single day. Consolidating these multiple parcels into a single, smart locker is another possible optimization. This would mean that the customer would have to make only one trip to the smart locker to pick up all their parcels and that the shipments would use only a single locker.

The smart locker delivery company needs to take into account the size of the smart locker, the number of packages that need to be delivered, and the time it will take to consolidate the packages into a single, smart locker. In some cases, delivering each parcel to a separate smart locker may be more efficient.

This decision can be made by using artificial intelligence to predict how long it will take to consolidate the parcels and compare this with the time it would take to deliver each parcel separately. By making this decision in real-time, smart locker delivery companies can optimize their resources and increase customer satisfaction.

Determining the Location of Parcel Lockers

All of the above solutions can be applied daily to improve last-mile logistics. But there is another important challenge that can be solved with machine learning: where should parcel lockers be located? As the smart parcel lockers market grows, the need for new locker locations appears.

The location of the parcel locker affects the number of customers that will use the locker, the number of parcels that can be delivered, and the km that need to be traveled. It is therefore important to choose the optimal location for a smart parcel locker.

Historical data need to be taken into account when choosing the new location, with available data points such as:

  • identifying popular parcel lockers,
  • customer density,
  • usage and filling rates of nearby lockers,
  • traffic patterns,
  • distance to retail stores,
  • area type - closeness to residential and commercial buildings,
  • distance to warehouse or distribution center,
  • distance to other parcel lockers.

By applying data analysis to these factors with artificial intelligence to find the optimal location for lockers within the smart locker system. By choosing the right spot, delivery companies can meet customer demands and deliver more parcels.

Adding Value to Smart Locker System

Smart parcel locker delivery is not only more cost-effective and convenient than home delivery, but it is also more eco-friendly. Thanks to advances in artificial intelligence and machine learning, we can now optimize the route and schedule and apply predictive analysis to accurately forecast delivery time, pickup times, or capacity rates for smart parcel lockers. This makes sure that packages are delivered on time and without issue.

If you’re looking for a more efficient way to manage your logistics services and business processes, look no further than artificial intelligence. With the help of artificial intelligence and big data analytics, you can improve your locker system and make it even more convenient for your customers. Thanks to route optimization, smart scheduling, and predictions, you can ensure that your smart parcel locker system runs at its best. So why not give AI a try? You might be surprised at how much it can help you improve your smart parcel locker system. Contact us today and discuss various machine learning solutions with our AI experts.

About the author

Dorota Owczarek

Dorota Owczarek

AI Product Lead & Design Thinking Facilitator

Linkedin profile Twitter

With over ten years of professional experience in designing and developing software, Dorota is quick to recognize the best ways to serve users and stakeholders by shaping strategies and ensuring their execution by working closely with engineering and design teams.
She acts as a Product Leader, covering the ongoing AI agile development processes and operationalizing AI throughout the business.

Would you like to discuss
AI opportunities in your
business?

Let us know and Mateusz
will arrange a call with our experts.

Thanks for the message!

We'll do our best to get back to you
as soon as possible.

This article is a part of

AI in Logistics
20 articles

AI in Logistics

Artificial Intelligence is becoming an essential element of Logistics and Supply Chain Management, where it offers many benefits to companies willing to adopt emerging technologies. AI can change how companies operate by providing applications that streamline planning, procurement, manufacturing, warehousing, distribution, transportation, and sales.

Follow our article series to find out the applications of AI in logistics and how this tech benefits the whole supply chain operations.

check it out

More articles

Logistics

Insights on practical AI applications just one click away

Sign up for our newsletter and don't miss out on the latest insights, trends and innovations from this sector.

Done!

Thanks for joining the newsletter

Check your inbox for the confirmation email & enjoy the read!

Find us on

Need help with implementing AI in your business?

Let's talk blue circle

This site uses cookies for analytical purposes.

Accept Privacy Policy

In the interests of your safety and to implement the principle of lawful, reliable and transparent processing of your personal data when using our services, we developed this document called the Privacy Policy. This document regulates the processing and protection of Users’ personal data in connection with their use of the Website and has been prepared by Nexocode.

To ensure the protection of Users' personal data, Nexocode applies appropriate organizational and technical solutions to prevent privacy breaches. Nexocode implements measures to ensure security at the level which ensures compliance with applicable Polish and European laws such as:

  1. Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) (published in the Official Journal of the European Union L 119, p 1); Act of 10 May 2018 on personal data protection (published in the Journal of Laws of 2018, item 1000);
  2. Act of 18 July 2002 on providing services by electronic means;
  3. Telecommunications Law of 16 July 2004.

The Website is secured by the SSL protocol, which provides secure data transmission on the Internet.

1. Definitions

  1. User – a person that uses the Website, i.e. a natural person with full legal capacity, a legal person, or an organizational unit which is not a legal person to which specific provisions grant legal capacity.
  2. Nexocode – NEXOCODE sp. z o.o. with its registered office in Kraków, ul. Wadowicka 7, 30-347 Kraków, entered into the Register of Entrepreneurs of the National Court Register kept by the District Court for Kraków-Śródmieście in Kraków, 11th Commercial Department of the National Court Register, under the KRS number: 0000686992, NIP: 6762533324.
  3. Website – website run by Nexocode, at the URL: nexocode.com whose content is available to authorized persons.
  4. Cookies – small files saved by the server on the User's computer, which the server can read when when the website is accessed from the computer.
  5. SSL protocol – a special standard for transmitting data on the Internet which unlike ordinary methods of data transmission encrypts data transmission.
  6. System log – the information that the User's computer transmits to the server which may contain various data (e.g. the user’s IP number), allowing to determine the approximate location where the connection came from.
  7. IP address – individual number which is usually assigned to every computer connected to the Internet. The IP number can be permanently associated with the computer (static) or assigned to a given connection (dynamic).
  8. GDPR – Regulation 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of individuals regarding the processing of personal data and onthe free transmission of such data, repealing Directive 95/46 / EC (General Data Protection Regulation).
  9. Personal data – information about an identified or identifiable natural person ("data subject"). An identifiable natural person is a person who can be directly or indirectly identified, in particular on the basis of identifiers such as name, identification number, location data, online identifiers or one or more specific factors determining the physical, physiological, genetic, mental, economic, cultural or social identity of a natural person.
  10. Processing – any operations performed on personal data, such as collecting, recording, storing, developing, modifying, sharing, and deleting, especially when performed in IT systems.

2. Cookies

The Website is secured by the SSL protocol, which provides secure data transmission on the Internet. The Website, in accordance with art. 173 of the Telecommunications Act of 16 July 2004 of the Republic of Poland, uses Cookies, i.e. data, in particular text files, stored on the User's end device.
Cookies are used to:

  1. improve user experience and facilitate navigation on the site;
  2. help to identify returning Users who access the website using the device on which Cookies were saved;
  3. creating statistics which help to understand how the Users use websites, which allows to improve their structure and content;
  4. adjusting the content of the Website pages to specific User’s preferences and optimizing the websites website experience to the each User's individual needs.

Cookies usually contain the name of the website from which they originate, their storage time on the end device and a unique number. On our Website, we use the following types of Cookies:

  • "Session" – cookie files stored on the User's end device until the Uses logs out, leaves the website or turns off the web browser;
  • "Persistent" – cookie files stored on the User's end device for the time specified in the Cookie file parameters or until they are deleted by the User;
  • "Performance" – cookies used specifically for gathering data on how visitors use a website to measure the performance of a website;
  • "Strictly necessary" – essential for browsing the website and using its features, such as accessing secure areas of the site;
  • "Functional" – cookies enabling remembering the settings selected by the User and personalizing the User interface;
  • "First-party" – cookies stored by the Website;
  • "Third-party" – cookies derived from a website other than the Website;
  • "Facebook cookies" – You should read Facebook cookies policy: www.facebook.com
  • "Other Google cookies" – Refer to Google cookie policy: google.com

3. How System Logs work on the Website

User's activity on the Website, including the User’s Personal Data, is recorded in System Logs. The information collected in the Logs is processed primarily for purposes related to the provision of services, i.e. for the purposes of:

  • analytics – to improve the quality of services provided by us as part of the Website and adapt its functionalities to the needs of the Users. The legal basis for processing in this case is the legitimate interest of Nexocode consisting in analyzing Users' activities and their preferences;
  • fraud detection, identification and countering threats to stability and correct operation of the Website.

4. Cookie mechanism on the Website

Our site uses basic cookies that facilitate the use of its resources. Cookies contain useful information and are stored on the User's computer – our server can read them when connecting to this computer again. Most web browsers allow cookies to be stored on the User's end device by default. Each User can change their Cookie settings in the web browser settings menu: Google ChromeOpen the menu (click the three-dot icon in the upper right corner), Settings > Advanced. In the "Privacy and security" section, click the Content Settings button. In the "Cookies and site date" section you can change the following Cookie settings:

  • Deleting cookies,
  • Blocking cookies by default,
  • Default permission for cookies,
  • Saving Cookies and website data by default and clearing them when the browser is closed,
  • Specifying exceptions for Cookies for specific websites or domains

Internet Explorer 6.0 and 7.0
From the browser menu (upper right corner): Tools > Internet Options > Privacy, click the Sites button. Use the slider to set the desired level, confirm the change with the OK button.

Mozilla Firefox
browser menu: Tools > Options > Privacy and security. Activate the “Custom” field. From there, you can check a relevant field to decide whether or not to accept cookies.

Opera
Open the browser’s settings menu: Go to the Advanced section > Site Settings > Cookies and site data. From there, adjust the setting: Allow sites to save and read cookie data

Safari
In the Safari drop-down menu, select Preferences and click the Security icon.From there, select the desired security level in the "Accept cookies" area.

Disabling Cookies in your browser does not deprive you of access to the resources of the Website. Web browsers, by default, allow storing Cookies on the User's end device. Website Users can freely adjust cookie settings. The web browser allows you to delete cookies. It is also possible to automatically block cookies. Detailed information on this subject is provided in the help or documentation of the specific web browser used by the User. The User can decide not to receive Cookies by changing browser settings. However, disabling Cookies necessary for authentication, security or remembering User preferences may impact user experience, or even make the Website unusable.

5. Additional information

External links may be placed on the Website enabling Users to directly reach other website. Also, while using the Website, cookies may also be placed on the User’s device from other entities, in particular from third parties such as Google, in order to enable the use the functionalities of the Website integrated with these third parties. Each of such providers sets out the rules for the use of cookies in their privacy policy, so for security reasons we recommend that you read the privacy policy document before using these pages. We reserve the right to change this privacy policy at any time by publishing an updated version on our Website. After making the change, the privacy policy will be published on the page with a new date. For more information on the conditions of providing services, in particular the rules of using the Website, contracting, as well as the conditions of accessing content and using the Website, please refer to the the Website’s Terms and Conditions.

Nexocode Team