What’s the difference between machine-to-machine communications and the Internet of Things?

The Internet of Things (IoT) is a big thing. Literally. It’s everything that is connected over the Internet.

The industrial flavor of IoT includes machines that will be able to share data and talk to one another. But there’s already a term for that, isn’t there? It’s called machine-to-machine (M2M) communications. Are these the same? Is the industrial IoT just a fancy new name for something that already existed? One thing is for sure: the IoT has a very global presence. We asked this international group of heavy hitters in the IoT domain to offer their perspectives on the differences and the similarities between M2M communications and the IoT. Learn more about IoT and what it means to manufacturing at Smart Industry.

Click here to meet the panel of experts.

What are the primary differences between machine-to-machine (M2M) communication and the Internet of Things (IoT)?

Vihang Sapale: M2M communication is a subset of IoT. We can consider a simple example. We have lot of sensors inside our car, which communicate with a microcontroller, and the microcontroller is taking certain actions based on the sensor inputs. Suppose we start the vehicle. One of the sensors would detect that engine has started, and the microcontroller in turn would activate the central locking system. We already have hundreds of such M2M-communication-based systems around us, such as washing machines and refrigerators. IoT does a lot of value addition to M2M communications.

In the case of IoT ecology, it involves sending this data over the cloud and performing various functionalities, such as analytics on the data, machine learning, remote monitoring and control. With reference to the automobile example, let’s consider that a manufacturer sends sensor data to the cloud and then can run analytics on the software and figure out vital information. Every year, hundreds of thousands of recalls are made by automobile companies to rectify issues they’ve identified in post-production, in order to take care of any liability arising in case of accidents. In this case, an automobile company can integrate its existing M2M platform with IoT to identify these issues at a very early stage and save millions of dollars in recalls. Also, the data can be analyzed for future trends or driving patterns to come up with radically improved products in future iterations.

Martin Harnevie: M2M is concerned with communication between machines, not necessarily connected to the Internet. M2M has traditionally also focused more on SIM-based cellular connectivity, whereas IoT is network-agnostic. Moreover, M2M has traditionally been concerned with single-use applications, which are closed to the outer world both in terms of its data and its application interfaces. In its original meaning, IoT was also about devices talking with other devices. IoT was the Internet of Things, implying the Internet between things, to which applications run on the Internet—for instance, Web applications—added substantial new value by making it easy to view and control them, regardless of location, as well as allowing shared services and shared data.

Also read: What is the danger of taking M2M communications to the Internet of Things?

Simplistically, one could then say that IoT = M2M + Internet. Over time, however, the IoT term has broadened significantly. IoT now encompasses all kinds of devices connected to the Internet, from industrial machines, smart home devices, wearables and environmental sensors to actuators and machinery that control the physical world. Thing-to-thing communication is no longer a central part of the IoT definition. Instead, as long as devices connected to the Internet feed data to Internet-centric applications and databases—the cloud—it is IoT. Currently, there are almost 10 billion devices connected to the Internet. This is expected to reach 50 billion devices within the coming decade. Together with advances in Web programming and new shared services and shared data business models, IoT is expected to have a disruptive impact on a global economical level.

Dwayne Dixon: There are a lot of different factors and subtleties that come into play here, and there is not universal agreement. I happen to differentiate them by thinking of M2M as communication between dedicated systems on a point-to-point basis. IoT on the other hand consists of sensors of many kinds from many manufacturers that connect in a horizontal way.

Glenn Vassallo: From my perspective it is cloud, and in particular the public cloud, that is primary difference. Access to the public cloud means access to powerful analytic options that would be extremely costly and difficult to implement in a local environment. Added to this is rapid implementation and flexibility, which all lead to value being extracted from solutions in significantly compressed time frames.

Peter Waher: The Internet of Things is what we get when we connect things that are not operated by humans to the Internet. It might seem obvious, but it has four distinct parts, which imply specific considerations that have to be made. First, “connection” relates to the study of communication protocols. The second part, “things,” relates to the study of sensors, actuators and controllers, among other things. The “non-operation by humans” is the third part, and it relates to provisioning and decision making. Fourth, the “Internet” relates to security, including identities, authentication and authorization, but also to interoperability.

While M2M solutions focus on connection and things, they often have very simplistic or non-existent solutions for non-operation by humans and the Internet. Their solution to non-operation by humans might be to allow all operations—if a controller receives a valid control command, it executes it without questioning the command. And the solution to the Internet might be to completely shut out third parties from their solutions using technologies such as encryption or virtual private networks (VPNs). But a true IoT solution must have greater visions than that. It must be able to interoperate with third parties, but in a secure and controlled fashion. This implies that the provisioning aspect must be much more developed, where each device is able to make the correct security decisions, without impeding functionality of the overall system. It must also work without the guidance of human operators. When third parties integrate, it should not require the operator of the IoT infrastructure or systems to make manual configurations. Only the third party and the owner of the corresponding things should be affected. It also means that the Internet part is much more evolved. Instead of shutting out third-party communication, such communication is permitted.

This, together with a vision of global scalability on the Internet, requires that architectures for federated authentication and authorization must be in place. This will allow interested parties to be correctly identified and security decisions to be correctly made. Interoperability also requires open standardized communication protocols and data models to be used, instead of proprietary formats, which might be sufficient in M2M solutions.

Francisco Maroto: Traditional M2M solutions typically rely on point-to-point communications using embedded hardware modules and either cellular or wired networks. In contrast, IoT solutions rely on IP-based networks to interface device data to a cloud or middleware platform. The difference between IoT and M2M is that IoT apart from the device-to-device criteria also makes use of Internet/cloud-based applications to not only effect the device-to-device communication, but also to make the information exchanged accessible using various platforms and/or open standards for other devices and also humans to partake in this exchange.

Generally, M2M does not do this, and, even if M2M sometimes also make use of the Internet, it is a proprietary application without any openings to other participants in the exchange of information. Traditionally, M2M has so far enabled remote management, operation of equipment/devices through cellular point-to-point connections. IoT/IoE/industrial Internet companies are new, aiming to target business transformations, which is good, but the broader canvas requires many other problems to be solved, such as privacy of data, ownership of data and security in connected cyber-physical systems.

Nicola De Carne: There is a significant difference between M2M and IoT, and there is also a lot of confusion and abuse of marketing terminology. M2M is a legacy solution to connect substantial machines to each other without the need to have a connection to the Internet and in a close system. It means that M2M could be more secure than IoT but also less scalable and flexible from both architectural and application points of view. There are a lot of telecommunications companies that are declaring to already have significant use case in IoT, but actually they are doing SIM-based M2M. IoT is the opportunity to connect the physical world to the digital world, to connect machines to the processes and to people that can leverage in a easy way the produced data to improve the efficiency of the machines and, in case of Industrial IoT, the entire production plant.

With an IoT architecture, you have a native IP communication layer, and you can easily expand the number of connected machines and add applications in the cloud or distributed at the edge. On the other hand, to start an IoT architecture in a plant you may have a significant impact on the process workflow, and sometimes it could be a barrier for the acceleration of applications in the market. Finally the IoT is not a revolution but an evolution. We started with telemetry transmission from the moon, then the PLC and now the IoT on IP. In each, we needed time to get the best results from new and emerging technologies.

Anshul Saxena: M2M is a subset of IoT. If we consider an atom, then M2M is the nucleus and IoT are the electrons of the atom. M2M is when machines talk to each other. It could be point-to-point interaction or interconnection of machines via the Internet. The former is M2M in its basic terms, and the latter is IoT. To expand the differences between M2M and IoT in today’s context, here are some other observations. M2M is peer-to-peer connectivity via Bluetooth, Zigbee or any other such protocol. IoT is M2M connectivity via cloud, involving Internet protocols. The applicability is in analytics and application layer abstraction. M2M is without human intervention. IoT may or may not be without human intervention. So, when it has human interaction, it technically is not a superset of M2M, as there is principally a violation of M2M principle.

Samuel Bucholtz: IoT is often more focused on the API, while M2M is focused on the data format, but the differences are rapidly fading away.

Jonathan Pollet: Machine-to-machine communications technically means communications between devices of the same type. This usually covers embedded devices communicating with other embedded devices. This category of communication is similar to Internet of Things; however, in the case of M2M communications, there is not an expectation that these M2M communications would communicate out to the Internet. The best example of M2M communications would be in the case of manufacturing systems whereby programmable logic controllers communicate with other controllers, conveyor belts, pumps, blowers and other plant control equipment at the same network layer to control or monitor the production of a product. M2M communications usually do not extend outside of their specific network layer.

Anand Gijare: IoT is beyond machine-to-machine communication and provides advanced connectivity of devices, systems and services. It covers a variety of protocols, domains and applications. IoT is the next step in automation in almost every field, including home automation, vehicle automation and automation of healthcare devices over IP networks.

What are the primary similarities between machine-to-machine (M2M) communication and the Internet of Things (IoT)?

Vihang Sapale: Both require reporting the machine data to a remote server or the user. Data reliability and security are prime concerns. And both technologies can provide real-time authenticated data to a single user or many users.

Jonathan Pollet: The primary similarity between M2M and IoT communications is that both cover communications with small embedded devices, machines or appliances.

Anand Gijare: Yes, connectivity or communication with each other is the primary similarity between M2M and IoT.

Samuel Bucholtz: Every M2M or IoT is an island. They all stand alone. Trust must be established and not assumed. A poetic way of saying that in the modern network environment is that every device must be responsible for its own security and must assume that the external systems it depends on or the external entities it communicates with will be breached.

Anshul Saxena: Both M2M and IoT refer to connectivity between two or more entities, and both have some protocol-based communication. M2M is usually a subset of IoT. Hence, whatever is M2M is also IoT.

Dwayne Dixon: Both use similar types of messaging, such as JSON, XML or other serialized data messages.

Francisco Maroto: M2M and IoT share this key feature: a device exchanges information with another device. In other words, both terms mean strictly device-to-device exchange of information. To me IoT solutions are facilitated by a collection of evolutionary technologies, such as mobile, cloud or sensors, and including those that have existed in the M2M world for several decades.

Martin Harnevie: Both M2M and IoT were originally about devices talking with other devices. This is implied in both terms. Typically, an M2M system has sensors, such as environmental sensors, position sensors, tilt sensors or proximity sensors, as well as devices and machines that perform functions: from production machinery, HVAC and door openers to industrial robots, vehicles and feedlot stations in farming. An M2M system also has an application that is configured to certain values, which can be part of an enterprise resource planning system, manufacturing execution system, production monitoring system or supply chain management system. In its most rudimentary form, IoT does the same, even though we talk about things in general and not only machines.

Main image courtesy of jscreationzs at FreeDigitalPhotos.net