Notre revue de presse de la semaine passée : du 15 au 26 novembre 2021.
The IoT Revolution: challenges and opportunities
Every once in a while, new technology appears with the potential to shift the entire social fabric of society. Lately, we have been hearing a lot about the “Internet of Things” (IoT), and wondering what to make of it.
According to the Research firm IDC, the IoT will become a $1.46 trillion international market by 2020, up from $700 billion last year.
The Internet of Things (IoT) generally refers to the extension of computing and network capabilities to devices and sensors that are not considered as computers, enabling them to make machine-to-machine interactions with minimal or no human input.
This technology is economically compelling, as the drive towards cost reduction for provisioning, energy savings, value added services and efficiency of management and use (automation) of equipment and infrastructure continues. Socially, this is already creating new services that were not possible before.
The heating and cooling system in your house could take cues from the motion sensors in the house, or receive commands from your smartphone, that help to reduce heating or cooling efforts when the house is empty. We can imagine now the smart assistant software running on our smartphones, sending notifications to the house, which activates the heating or cooling start when one is thirty minutes from home. Similarly, a diagnostic machine or medical device in a doctor’s office or lab hooked up to a private cloud can trigger a technical support call, or the shipment of new consumables from the supplier when required. Other existing examples that have been implemented in some form include having a Manufacturing Execution System trigger the ERP (Enterprise Resource Planning) to order new parts for a manufacturing plant, or the reading of electricity consumption and automatic billing of customers. It is not even a far-fetched idea to have your refrigerator trigger a service call or order milk automatically from a nearby shop for delivery when you use the last drop in your morning cereal. The only limit is your imagination.
Some of the current leaders in this area are companies such as Amazon, Bosch, Axeda, Google, Siemens, Honeywell, Cisco, GE, Dell, Hitachi Data Systems, Huawei, Intel, IBM, Microsoft, Oracle and Samsung. Axeda offers a complete software stack for certain areas of application; though it is still too costly for generic adoption. For Industrial Internet of Things (IIoT), National Instrument offers a specialized stacks that includes the Software and Hardware (LabVIEW, Lookout, PXI and embedded controllers and accessories, some with FPGA). One company not mentioned above, but which has the potential to transform IoT, is Apple. Their hardware, software stack and cloud services allow them to create a full consumer IoT anytime. Apple’s HomeKit is a timid start as it is aimed only at the residential home market. The HomeKit is a framework for communicating with and controlling connected accessories in a user’s home.
Though many of us are already participating in the IoT revolution without even knowing it by buying products that support these features. The most popular products in this category are internet-enabled television sets, cell phones and tablets. Once configured, these interact daily with cloud services and service providers’ servers without any intervention from us.
IoT devices with limited functionality have been around for at least a decade. What has changed recently is the ubiquity of connectivity options (WIFI, 3G, and Bluetooth etc.), cloud services and analytics, which are great enablers for IoT. The Cloud provides a platform for hosting intelligent software, networking a large number of IoT devices and provisioning them with a large amount of data. This enables smart decisions to be made without human intervention.
However, there are still some current challenges limiting the adoption of IoT:
Security vulnerabilities (privacy, sabotage, denial of service): Regular hacking of high-profile targets keeps this danger constantly in the back of our minds. Obviously, the consequences of sabotage and denial of service could be far more serious than a compromise of privacy. Changing the mix ratio of disinfectants at a water treatment plant or stopping the cooling system at a nuclear power plant could potentially place a whole city in immediate danger.
Regulatory and legal issues: This applies mainly to medical devices, banking, insurance, infrastructure equipment, manufacturing equipment, and in particular, pharmaceutical and food related equipment. Today, this mean complying with laws such as CFR 21 part 11, HIPAA, Directive 95/46/EC and GAMP 5. Etc. This adds to the time and cost needed to bring these products onto the market.
Determinism of the network: This is important for almost all areas where IoT can be used, such as in control applications, security, manufacturing, transport, general infrastructure, and medical devices. The use of the cloud currently imposes a delay of about 200 milliseconds or more. This is fine for most applications, but not for security or other applications that require a rapid, almost immediate, response. A trigger from a security monitoring system received five seconds later could be too late.
Lack of a common architecture and standardization: Continuous fragmentation in the implementation of IoT will decrease the value and increase the cost to the end users. Currently, aside of the products mentioned above, there are also Google’s Brillo and Weave, AllJoyn, Higgns, to name but a few. Most of these products target very specific sectors. Some the causes of this fragmentation are security and privacy fears (privacy through obfuscation and the fear of “not invented here”), jostling for market dominance, trying to avoid issues with competitors’ intellectual property, and the current lack of clear leadership in this area.
Scalability: This is currently not much of an issue, but it is bound to become an issue mainly in relations to generic consumer cloud as the number of devices in operation rises. This will increase the data bandwidth needed and the time needed for verifying transactions.
Limitations of the available sensors: Fundamental sensor types, such as temperature, light, motion, sound, color, radar, laser scanner, echography and x-ray, are already quite performant. Furthermore, recent advances in microelectronics, coupled with advances in solid state sensors, will make the bare sensors less of an issue in the future. The challenge will be in making them more discriminating in crowded, noisy and more complex environments. The application of algorithms that are similar to fuzzy logic promises to make this less of an issue in the future.
Dense and durable off-grid power sources: While Ethernet, WIFI, 3G and Bluetooth have been able to solve most connectivity issues by accommodating the various devices’ form factors, the limitations of battery life still remain. Most smartphones still need to be charged every day, and most sensors still need regular battery changes or connection to the grid. It would make a difference if power could be broadcasted wirelessly to such devices from a distance, or if power sources that can last for at least a year can be integrated into the sensors.
Recent initiatives such as “Open Interconnect Consortium”, “The AllSeen Alliance” and the proposal from IBM (Adept) for the use of modified Bitcoin block chains in IoT. A blockchain is a distributed public database that maintains a continuously-growing list of data transactions, protected against tampering and revision. It grows linearly and chronologically as new completed blocks are added to it (see Baris Yeldiren’s article on “Blockchains – the next big thing”). Blockchains hold promises in helping to resolve most of the above problems and in alleviating integration challenges. Bitcoin blockchains take a lot of computing power to generate. IBM proposes to increase speed and scalability by partially replacing the “proof of work” part of the block by a “proof of stake”, which require far less computations.
When it comes down to it, the IoT is around to stay. Most of the technical problems will be resolved within the next few years; but it is the company that will come up with a generic framework and compelling business model on how to implement and monetize IoT that will possess the goose that lays the golden eggs.
Image credit: jeferrb via Pixabay, cc0 Public Domain