Assessing Buzzwords: X “as a Service” (XaaS)

Owning a car used to be a fundamental part of easy transportation access. Today, one can summon a car with a phone app, and a car will be available. You need not worry about maintenance, upkeep, damages, insurance, or the like.

This is possible by services like Uber and Lyft, called transportation-as-a-service (TaaS). The service model is expanding into more areas, including the industrial space. Products can be rented for a limited time and returned when not required. Such offerings are commonly termed “(Blank) as a service” or “X (anything) as a service” (XaaS).

 

The Rise of “as a Service” Model

The business model of “as a Service” has been around what seems like forever. For example, Blockbuster used to rent movie cassettes and CDs at a fraction of the cost required to buy the movie outright. But the term “as a Service” became popular with the rise of the internet and network-based economy.

The internet is a collection of computers that talk to each other over a communication channel. The websites and applications have to be hosted on a computer called a server. This server should be available around the clock for undisrupted access to the website. Individuals are unable to manage the quality of infrastructure needed to make servers highly available. 

 

Figure 1. Representation of a server room and IIoT. 

 

Service providers offered server infrastructure to anyone who wants to host a website. They manage all the necessary physical infrastructure, and the user can manage the platform and software running on the infrastructure. 

The availability of virtualization technology enabled such service providers to divide powerful physical infrastructure to different users. This evolved with the advent of Amazon Web Services (AWS) and other cloud service providers offering more services with their cloud platforms. They offer infrastructure, platforms, storage, computing, software, etc., as a service so anyone worldwide can use it at a minimal cost.

The term “as a Service” became popular with the widespread adoption of SaaS or Software as a Service. SaaS providers use the platform/infrastructure other cloud service providers offer to build software applications other businesses can subscribe to “as a Service.” Later, Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) monikers were introduced even though the business models existed before SaaS became popular. 

 

Types of “as a Service” Products

Many services have adopted the “as a Service” business model. Initially, it was limited to computing infrastructure and services. Today, physical entities used in everyday service can be used in an “as a Service” model.

 

Infrastructure/Platform as a Service (IaaS and PaaS)

Computing hardware, platforms, and other hardware can be similarly rented. The users need to pay only for services they use. 

When only hardware infrastructure is involved, it is termed Infrastructure as a Service (IaaS). When hardware along with the platform is involved, it is termed Platform as a Service (PaaS). Google Cloud, AWS, Microsoft Azure, etc., are examples of IaaS/PaaS providers. 

 

Figure 2. The three-layer factory network model pyramid. Image used courtesy of Analog Devices

 

The infrastructure may be charged at a fixed rate per month as part of a service contract. Service providers can offer sensors, computers, and other network devices as part of such service contracts. IaaS/PaaS service providers also employ a pay-as-you-go structure where customers pay only for the resources they use rather than a fixed rate. 

 

Software as a Service (SaaS)

Businesses can use various software on a subscription basis, meaning they can unsubscribe when the software’s usefulness expires. Such services are generally built on top of IaaS or PaaS. In most cases, SaaS providers offer free trials for prospective customers. The purchase decision is made after a satisfactory trial.

Some examples of SaaS include the following.

• ThinkIQ offers data functionalities to manufacturers to implement smart manufacturing solutions.
• Neurala is a SaaS product designed to help customers refine creating custom vision artificial intelligence (AI) solutions.

 

Other types of SaaS include enterprise resource planning (ERP) and warehouse management systems (WMS). Similarly, these products often coincide with IaaS or PaaS because they work with an existing platform. 

 

Analytics as a Service (AaaS)

Businesses can offer data analytics “as a Service.” Big data has become a critical component to gain a competitive advantage, but a manufacturing organization may not have a core competency in analytics. Modern analytics require handling big data and predictions. 

Building up analytics know-how can take up significant capital, resources, and time. A stabler strategy is to outsource analytics to organizations with strength in analytics. This way, the manufacturing firm can continue working with its core competency (i.e., manufacturing). Analytics as a service provides the data the facility needs to make decisions. 

One such decision is predictive maintenance. In this practice, engineers use gathered data to predict when and how a machine may fail. It is one of the biggest-growing needs in industrial analytics and data science.

 

Robot as a Service (RaaS)

Many types of robots can be used in an “as a Service” model. Such services are useful in two instances, among others.

• When there is a specific one-time use for a particular robot
• Unscheduled downtime for a robot used in a facility

 

Drones used for visual inspection may only be used only once, so they can be rented from service providers for a brief time and returned once the use is over. Similarly, automatic guided vehicles (AGVs), sorting robots, autonomous mobile robots (AMRs), or pick and place robots can be used in the service model for one-time use or as a replacement. 

 

Figure 3. The Rapid Machine Operator (RMO) comes with integrated cloud software. Image used courtesy of Rapid Robotics

 

Another area of RaaS includes SaaS capabilities. Robot programming can be complicated, so some robotics companies provide cloud services along with their robots. For instance, Rapid Robotics provides a yearly subscription that includes a robot as well as cloud-based machine learning.

 

The Cost of “as a Service” Products

The advantage of using “as a Service” infrastructure is the potentially incredible associated cost savings. One of the primary costs to build the necessary infrastructure is the capital expenditure required at the start of the project. Using an established infrastructure brings down this cost. The capital expenditure will be divided among all the parties using the “as a Service” infrastructure across the break-even period. 

Similarly, the maintenance cost of the infrastructure will also be handled by the cloud service provider. Service model users should not have to worry about what goes on under the hood. They just need to plug and play the “as a Service” solution and pay only for the services they use. 

Cybersecurity is another concern that should be handled by the “as a Service” provider. Today, it is critical to have strong cyber resilience in place to thwart malicious attacks. Small companies hiring a cybersecurity team may have a very high cost attached. Pooling the resources for many customers, “as a Service” providers can offer better cybersecurity for a lower cost. This may be a fraction of the cost they would have encountered if they had to perform cybersecurity themselves.

The cost benefits are also evident when the infrastructure is no longer needed or needs upgrading. If a company had built the infrastructure on its own, it would have to find ways to dispose of the infrastructure. It will be a non-performing asset on the balance sheet and can only be removed by suffering a considerable drawdown from the original cost. But stopping infrastructure from “as a Service” providers is as easy as stopping a magazine subscription.

There are few instances where building their own infrastructure is beneficial. Most of the time, it would be required by large companies like Facebook with global operations, and owning the infrastructure would be much more cost-efficient. Another case is where security and privacy concerns are so high that using an “as a Service” provider is risky, such as critical infrastructure.

The “as a Service” business model has evolved and is used across a broad range of industries and sectors. More products and services outside of industrial and transportation will likely start using this business model. This will be a beneficial relationship for the consumers and “as a Service” providers. Will the service model continue an upward trend, or will it soon crash?