Get Hazelcast

IoT Analytics and Telematics at the Largest Logistics Operation in the World


IoT Analytics and Telematics at the Largest Logistics Operation in the World


Santa’s global operation requires the speed and efficiency of a real-time enterprise.

Every year starting in late November until the new year, the holiday season is upon us, when families and friends join together for festive gatherings. Celebrating the holidays often includes spending time with the family and exchanging gifts. Sometimes the season can be stressful when it comes to buying presents. So imagine the hectic time for the largest logistics operation in the world.

Background

With Santa’s Nice List growing every year, he must maintain a streamlined, real-time enabled operation.

No matter which holiday (or holidays) you celebrate, you likely enjoy the fun (and madness) of trying to get the right gift for your family and friends. That’s a big part of the season traditions.

One well-known contributor to holiday cheer is Santa’s Workshop at the North Pole. Its operations include not only Santa Claus’ delivery efforts on Christmas Eve but also all the planning and coordination by his team of elves. It is like a delivery service, a supply chain service, a manufacturer, a retailer, a warehouse, and an information services business all rolled into one enterprise.

Santa’s global operation is obviously very time-sensitive and requires the speed and efficiency of a real-time enterprise.

Consider all the elements of Santa’s work—the legendary Naughty or Nice list, the “Dear Santa” letters from kids around the world, the photo opportunities at the local shopping mall, just to name a few. There is so much to track that everything needs to be extremely streamlined to minimize the potential for error. And in his business, errors mean unhappy kids, which is unacceptable to him.

The Time to Modernize Santa’s Global Logistics Operation Has Come

In the past, Santa had far fewer homes to visit and far fewer gifts to deliver. And during those early days, his team of elves could build all the toys. But now Santa’s Workshop works directly with toy manufacturers and retailers to help fulfill the growing list of presents to deliver. That adds a new level of coordination, which makes things more difficult.

So it’s not surprising that Santa runs the most extensive logistics operation on the planet, and it gets bigger each year. No matter what’s happening in the world, Santa still makes the magic happen. And while he makes it seem easy, it definitely is not.

Optimization via Telematics

To help optimize his shop, he relies heavily on data from various sources that his IT and analytics teams continuously analyze to gain important insights. One of the key data sources is telemetry data from his sleigh and reindeer, which is analyzed in real-time to make immediate route recommendations and to ensure his reindeer are flying with absolute focus and determination. This telematics use case has high throughput and low latency requirements in a real-time environment similar to the workloads that many commercial businesses run today, like in financial services, e-commerce, telecommunications, insurance, energy, etc.

Challenges

Santa is always looking for ways to spread more cheer in less time — by creating optimal routes, he can ensure he doesn’t run behind schedule.
Real-Time Means On-Time

Having primary operations based at the North Pole makes delivery especially hard since much of his “client” base is far away. And there are countless factors to consider when delivering presents on time. Snowstorms, rain, changes of address, kids staying up late, fireplaces remaining lit, last-second modifications to the Naughty/Nice list—it all must be accounted for in the real-time route planning analysis. Since all of these factors change and are not necessarily known in advance, Santa needs to be acutely aware of immediate events and act immediately.

Workloads Grow along with Expectations

Also, while the world’s population continues to grow, the number of hours in the day remains the same. More kids and more presents mean the workload grows at a crazy rate. Santa hopes that outsourcing the delivery work will be possible in the future, but he does not expect that to be an option any time soon. So he is continually looking at ways to do more work in less time. The highest speed when analyzing data is essential, as any delays may lead to missed deliveries (and missed opportunities for happiness). And in addition to spreading cheer, Santa has a reputation to uphold, and he can’t afford missteps that could seriously undermine his credibility.

“I know it’s not a competition, but I have the biggest logistical challenges in the world.” — Santa Claus

Even Santa Has a Budget

From an IT perspective, Christmas Eve factors represent a flood of data that must be analyzed within millisecond windows. This means the efficiency of Santa’s sleigh needs to be nearly perfect. He determined long ago that legacy systems could not keep up with the growing avalanche of data that enables him to plan his global trip. With the largest data set of any organization in the world, by far, he knew he needed to turn to newer technologies to help his operations. And while most people believe that Santa Claus has infinite resources for his work, he does not. Like most businesses today, he must be conscious about his expenditures on resources that are at the heart of his IT division. His requirements include high performance, extreme efficiency, and reliability.

Solution

An edge computing infrastructure allows Santa to monitor household fireplace useage, whether children are asleep, and if cookies have been set out yet.

Some folks claim that Santa was the first user of real-time route optimization systems, and considering his leadership in delivery scale, that makes sense. Today, more organizations are taking advantage of tracking connected vehicles to optimize their fleet performance, and they can certainly learn a lot from Santa’s experience.

Machine Learning-Based Route Optimization

The Workshop starts with collecting current weather data from a wide variety of public and commercially licensed sources to get a complete picture of opportunities and obstacles in his path. His machine learning (ML) initiative has plotted, and continually refines, how his sleigh behaves in different weather conditions. As he detects heavier storms that might inhibit his progress, the system will redirect him to calmer weather on the fly to ensure he does not face unnecessary delays. Because of insights generated by his system, some longer paths will take much less time because of the lesser interference from storms. And as new information arises, he can dynamically alter his route, so any changes in weather patterns can be accommodated in the system at millisecond (and even microsecond) speeds.

Edge Computing and Stream Processing

Santa also has a massive edge computing infrastructure that can collect household data (approved by parents, of course) that monitors the status of the households. He wants to make sure all kids are asleep. He wants to make sure the fireplaces are out so he can climb down the chimney. And if the fires are not yet out, he wants to know if there’s an alternative way to reach the Christmas tree. He even tracks whether households have had the chance to set their milk and cookies out. All of this information also contributes to planning the optimal routes in real-time.

The final factor that Santa considers is the potential for the distraction of his reindeer during his sleigh ride. He cannot visually monitor all nine reindeer while making deliveries, so he relies on individual Internet of Things (IoT) sensors. Similar to the notion of “distracted driver analysis” for vehicle fleets, this system collects many different data points many times per second to look for patterns that represent reindeer distraction, making sure their productivity is unhindered.

A Digital Integration Hub on the Hazelcast Platform

The Workshop deployed a digital integration hub that acts as a powerful data layer that collects data from many sources and prepares that data via parallelized, distributed computing functions, for delivery to other analytics systems. The overall solution is a very sophisticated environment with a high degree of modularity that helps the huge development team focus on their individual responsibilities while simplifying the coordination with other teams.

With the Hazelcast Platform, Santa can connect to all of his data sources and run computations with in-memory speeds. Stream processing capabilities are the core of its real-time processing, and it ingests, transforms, enriches, and aggregates data instantaneously. Both the in-memory data store and the stream processing engine work together to deliver a complete real-time solution that creates context around the comprehensive data set and takes action on the insights that are created. With its elastic architecture, Hazelcast clusters can grow as more data is collected, and then shrink in size during the non-holiday seasons.

“After centuries of waiting, a real-time technology solution finally exists.” — Santa Claus

The Workshop built a microservices architecture based on Kubernetes that lets separate teams of developers focus on a specific function. Microservices give them the flexibility to scale individual parts of the system, while also simplifying the development effort by building small and manageable apps that make up the overall system. They use Apache Kafka as a messaging bus to pass data between microservices as an efficient way to connect them together.

Since not all elf developers are Java experts, the .NET, C++, Go, Python, and JavaScript developers can use the respective Hazelcast client API to build their microservices. This is especially important for the elf data scientists who love Python, as they can deploy their Python-based machine learning models into a Hazelcast pipeline without any intervention from the core Java DevOps elves. The highly efficient Hazelcast Platform lets them get the highest throughput from their machine learning models since Hazelcast pipelines are optimized to take advantage of the available resources in a highly parallelized way.

Cloud Topologies

The Workshop uses a combination of Hazelcast software and Hazelcast Cloud, the fully managed service running the Hazelcast Platform. That’s because the Workshop deploys private cloud, hybrid cloud, and multi-cloud topologies to support the global operation. These separate deployments leverage the advantages of each site. For example, data that is created and stored at the North Pole is processed on-premises, while data collected at remote locations is typically assigned to the nearest public cloud vendor site to reduce latency across the network. This geo-distributed architecture is especially important since Santa covers so much of the world. The combined use of these topologies also helps with ensuring a highly resilient system that includes a solid disaster recovery strategy. With an unquestionable need for low recovery point objectives (RPO) and low recovery time objectives (RTO) in their business continuity strategy, Hazelcast has them covered. Hazelcast simplifies the synchronization of these separate deployments via the WAN Replication feature, and the Automatic Failover feature lets them switch to a secondary cluster should the initially connected cluster become inaccessible.

Results

With extreme high performance and the real-time capabilities that Hazelcast added to the IT team at Santa’s Workshop, Santa can continue delivering toys and gifts on time while spreading happiness each winter. All of this is possible despite the scaling challenges that he faces each year. He is able to fulfill the promise of “the right toy for the right kid and the right time.”

“I have more time to spend on what matters most.” — Santa Claus

He measures results and continues analysis, especially machine learning-based, to look for more opportunities for optimization, since his workload will not get lighter any time soon.

Santa continually accomplishes what appears to be impossible, mostly by thinking beyond the possible. Like many of today’s successful businesses, it’s not acceptable for him to be complacent with the status quo, especially when his business climate continues to become more challenging over time.

Kids worldwide will continue to enjoy Santa’s hard work and generosity. This is only possible by operating as a highly data-driven, real-time enterprise that turns to technologies like Hazelcast to power large-scale data processing and analytics.

Let's Get Started!

Contact us now to learn more about how our in-memory computing platform can help you leverage data in ways that immediately produce insight and actions.