IoT to IoAT: Internet of Autonomous Things devices provides solutions

Future Internet of Autonomous Things (IoAT) devices will utilize knowledge-enhanced electronic logic (KEEL) technology and may consume information from other devices or the cloud and participate in solutions they were never designed for.
By Tom Keeley, Compsim April 30, 2016

Figure 1: The Internet of Autonomous Things (IoAT) is designed to query information in the cloud to think about what to do next. Courtesy: CompsimThe next generation of Internet of Things (IoT) devices will deliver expertise and adaptive command and control, beyond just providing information for higher level processing. Knowledge-enhanced electronic logic (KEEL) technology will play a role in accelerating the delivery of these advanced capabilities into small, low-cost devices. The big data concept is that all of the connected devices will be producing information that is consumed by some higher-level system. Potentially, there is another view.

Given that some type of controller will control these devices, many of these devices can take on new responsibilities. Gartner Inc. predicts that there will be 6.4 billion connected objects in use by 2016 and 21 billion by 2020. Rather than just producing data, these devices can perhaps take on additional roles. In the future, these devices may consume information from other devices (and the cloud) and participate in solutions they were never designed for. 

Not just collecting data

The primary objective of collecting data usually is to make better decisions. Predictions using data can prepare systems for the future by detecting change, or exerting some kind of control. Using a broader description, one might suggest that data can be collected for the purpose of controlling behavior. If the organization doesn’t control systems’ behavior, another organization will collect that data and use it accordingly. The common perception is that as problems become more complex, large processing engines can only handle them, or that those problems require humans-in-the-loop to interpret the complex information sets.

Another common "big data" view is that much of the information processing will be accomplished by searching for patterns of data from large data sets that are collected over time. This view is that the data is in control, so we should let the data drive the solution. Well perhaps, or perhaps not?

Some who have been around for a while look at the IoT and see nothing that is really new. Distributed control and supervisory control and data acquisition (SCADA) products have been used in industrial automation for many years.

"Timesharing" and "cluster computing" are terms that have been associated with distributed computing. These terms are often used to define the technology-of-the-day that includes taking inputs, manipulating the data, and then distributing information to control actions or outputs.

The evolution of technology that has resulted in the IoT market has been driven by the commoditization and re-distribution of resources. Each time a shift takes place, some marketer will create a new name and claim the market.

The commoditization of processing power in microcontrollers, tied to low-cost development environments, has reduced the cost of processing information. It has also commoditized interconnectivity with consumer-based networks and protocols that have provided the infrastructure for new devices to participate in more complex applications. Additionally, microelectromechanical systems (MEMS)-based devices have reduced the cost of sensors and actuators.

Cloud-based control solutions

Figure 2: The Internet of Autonomous Things (IoAT) allows devices to provide actionable data back to the cloud. Courtesy: CompsimCloud-based control solutions help many organizations centralize their processing deployment, which may help them manage their distributed applications.

However, there is always a "but" in this type of situation. Sometimes things go wrong. Putting all the eggs in the "cloud basket" connected by webs of open networks exposes many organizations to new risks. The new market for network security, redundant communications, and encryption has recognized the risks and offered patches to protect the user from new problems.

If a system can be hacked, it will be hacked. And given that if a system can break, it will break. And when things go wrong, they will probably go wrong at an inconvenient time.

Consider this analogy: What if there was another Earth, "Earth1," that only had one human living on it? That single human could have a billion tentacles that are connected to billions of tools. Earth1 could operate just like Earth; possibly even better, because, hypothetically, the single brain of Earth1 could manage conflicts between its tentacles and its tools (see Figure 1). However, if Earth1 had a problem and lost some of its tentacles, those functions would be completely lost. And if Earth1 forgot how to process all the information, then it would be dead in its universe.

Compare this to our Earth with its billions of humans. Each human has a brain that has its own proprietary sensors and actuators (see Figure 2). Our Earth is not dependent on any single communication link. Furthermore, it is not dependent on any single human. Our Earth benefits from groups of humans working together and it benefits from humans redistributing themselves from one collection of humans to another to address different problems. It also benefits from the ability of individual humans to continue to operate when communication links are broken.

Figure 3: Distributed expertise allows devices within the microgrid to either cooperate or self-organize its data based on the situation. Courtesy: CompsimHumans (in general) can figure out what to do if one of their tools is broken or lost. Humans can also respond to damage to other humans, as well as to bigger strategic and tactical issues. Individual humans can focus attention on specific tasks, but they know when to look beyond their assigned tasks at the "bigger picture." Humans (as a population, or groups of populations) are able to adapt (see Figure 3).

Another characteristic of humans is their ability to dispense "expertise." Expertise is more than having an ability to follow rules. Expertise is the ability to deliver judgment and reasoning. Judgment and reasoning are what allow the human to balance alternatives. They understand when to re-allocate attention when unexpected things happen. Human judgment and reasoning are used to solve more complex, inter-related problems that have conflicting objectives (solving tactical short-term goals while still considering longer term strategic objectives). It is judgment and reasoning that help humans make relative (how much) decisions, as they balance risks and rewards. 

Current state of the IoT

The general concept is that IoT devices will be generators of information. However, maybe the edge devices should also be viewed as consumers of information (using our human model). There might also be an opportunity to expand more into the actuation role and do some things locally without depending on the computer in the sky to process information.

In this case, users might avoid some of the risks associated with propagation delays and widely distributed computing. If we look at opportunities for IoT devices to behave as autonomous (or semi-autonomous) devices and redefine IoT as the Internet of Autonomous Things (IoAT), then what might the market look like? 

IoAT: Internet of Autonomous Things

Since there are an abundance of information sources distributed across the face of the Earth, maybe there will be a growth in distributed tools. It is very likely that more intelligent actuators will be created.

Based on that, the following scenarios are likely:

  • Actuators will collaborate with local information sources (just like humans do). Microgrid infrastructures are starting to appear in the market.
  • Actuators will be consumers of information (as humans are).
  • Actuators will identify their own information sources beyond those directly attached (just as humans gather information from nearby information sources and use their own senses to drive their own decisions and actions).
  • Actuators will work together to address problems when they lose connectivity with their supervisor in the command hierarchy (just as humans operate with their peers in emergency situations).
  • More intelligent autonomous/semi-autonomous devices will be able to recognize unplanned situations and respond according to human-defined guidelines (just like humans follow rules of engagement and operational policies they have been provided in advance). This will enable the devices to address problems they have never encountered before.
  • When dealing with teams of devices working together, there will be devices that may encounter problems during operation. These self-organizing devices will be able to react in real time (adapt collectively) to change.

Even when these IoAT devices (operating as a local team) encounter problems they cannot address together, they may be able to transmit actionable intelligence up their information hierarchy (if they have a communication link). They will even be able to use their own embedded expertise to look for communication alternatives.

Think of these new IoAT devices as objects that can operate independently or as objects that can self-organize and operate as teams to solve problems they have never been programmed to specifically address before.

More mobile-like devices that have multiple tools attached will be able to address more problems, all without going to the cloud for advice.

One might speculate that personal security, home automation, health care, industrial automation, transportation, agriculture, and military applications will lead the way with IoAT devices.

One might also speculate that personal, security, and financial applications will lead the way in the form of personalized software agents.

What are the roadblocks?

There are still some challenges and hurdles because it remains in its infant stage.

The hobbyist market is exploding with remotely piloted aircraft. While many of these devices are still remotely controlled, many have significant processing power. A DIY drone has GPS, roll-pitch-yaw sensors, altimeter, compass, video feed, wireless network connections, battery sensor, torque sensors, and motor feedback logic. Platforms come in various sizes and demonstrate how much performance (with connectivity) is currently available in the commercial space (i.e. commoditized). So, processing power is not a roadblock.

Connectivity is also not a roadblock since the Internet has commoditized connectivity. Wi-Fi and Bluetooth will continue to add security to their information exchange. Peer-to-peer and point-to-point microgrid networking will open new doors. It is likely that more development will be made in this area. The consumer and hobbyist market will drive this. Locally connected devices will be able to handle various point-to-point, peer-to-peer, and broadcast messaging structures.

Also, consider the IoAT actuators. This will be a new market for many new players. These will be the tools, arms, and hands (end effectors) of the IoAT devices. Their basic function will be to respond to threats and opportunities. Initially they may be remotely controlled tools, but eventually they will be able to self-organize and act collectively (team). They will become the human assistants that are always alert, always doing their job, always ready to respond to new threats and to new opportunities. Rather than calling the availability of these devices roadblocks, they will be market opportunities.

One might suggest that their availability will create new opportunities because some problems will need support from multiple IoAT devices working together (so they can collectively solve problems). Looking at the opportunities will drive changes: Why not solve the problem at the source, rather than just providing information for someone else to worry about?

There has been another significant roadblock. There will be a requirement to be able to capture and package human-like judgment and reasoning skills that will execute in the small, low-cost IoAT devices. We are talking about packaging human-like expertise that will enable these devices to solve complex problems that have historically required humans directly in the loop to address.

However, we do not have to replicate the human brain in these devices. These are machines with some limited set of capabilities. We only have to enable the machines to have the ability to consider alternatives of how they can use their capabilities in different situations. We just have to give them the ability to think in a more abstract manner. We do not want them to have free will. We do not want them to decide who their master is.

Also, we do not want to give the devices a weapon and have it used against us. Again, they will remain just machines; but they will be more capable machines that can use their abilities to solve more problems than they were originally built for. We want these devices to be able to pursue broader goals.

Business drivers

Location is everything. Users deploying packaged sensors are occupying space. A machine in place is already occupying that space. By adding new capabilities in the same space, the user is delivering a better level of service. Users and companies that are able to package human-like judgment and reasoning with the solution will create new opportunities and use that capability for a more efficient operation. It cane also be used to add more safety to an enterprise on enhance security.

In a factory automation environment, the user can improve operational efficiency if the machines can monitor and adjust their own behavior without depending on simple, statistical, preventative, maintenance procedures or dedicated operators. If machines can monitor their own stress, age, wear and tear, and upcoming jobs, they can tell their supervisors when maintenance should be performed in advance of system failures. Or they can highlight significant risks that they see coming.

Business drivers for someone interested in delivering "expertise" through IoAT devices:

  • Provide new capabilities:
    • Take humans out of the control loop
    • Reduce human error by automating services
    • Insert expertise where humans cannot go (size/risk)
    • Allow humans to command their own army of devices (amplify a human’s capability)
    • Provide services that historically have required humans to analyze, thus making human expertise available continuously. 
  • Provide the ability to package judgment and reasoning into software applications and devices and make it:
    • Easy to use
    • Easy to learn
    • Easy to test
    • Easy to explain. 
  • Provide a solution that can be deployed in very low-cost devices without adding significantly to base costs.
  • Provide a safer environment:
    • Remove humans from risk
    • Reduce human-induced errors in judgment.
  • Provide 100% explainable and auditable behavior:
    • For safety critical systems
    • Because the user wants to know why things happen
    • So the user can focus on making systems better and better.

The market for IoT solutions will continue to evolve. IoAT devices will process information closer to the source to determine:

  • What does it all mean?
  • What can be done and how should it be done?
  • Who can help?
  • Who needs to be told what to do about a situation (providing actionable intelligence; not just data)?
  • What more is needed to address the situation (beyond what an individual device, can do)?

Delivery of expertise that is closer to the information source will allow minor problems to be addressed before they become major issues. This localized processing of information will isolate the user from many of the security issues that may exist with distant, centralized information processing.

Organizations will compete based on the packaged expertise and the adaptivity of the IoAT devices they offer to the market.

The tools and techniques that will allow these capabilities to be delivered exist today. All that is required is for alert organizations to move quickly to address the opportunities.

Tom Keeley is the founder of Compsim and the inventor of knowledge enhanced electronic technology. Edited by Joy Chang, digital project manager, CFE Media, jchang@cfemedia.com.

ONLINE extra

See related stories about the Internet of Things (IoT) linked below.

Want this article on your website? Click here to sign up for a free account in ContentStream® and make that happen.