Sensors in the era of cloud connectivity

History.com reports that the first industrial revolution (circa 18^th and 19^th century) saw the introduction of automation in the manufacturing process. The premise of making things better, faster and with greater volume hasn’t really changed much.

The fourth industrial revolution, also referred to as Industry 4.0, continues this trend albeit with more advanced tools and technologies that facilitate greater automation and in today’s information-dependent business models more accurate data exchanges.

Industry 4.0 has given rise to the development of “smart factories” where modular structures are the norm, as is the convergence of cyber-physical systems to enable better monitoring of physical processes.

A key technology that will see greater development is in the Internet of Things (IoT) – highly specialized devices that perform a simple task, but do it well, fast and cheaply – and combined with other innovations like artificial intelligence, virtual reality, advanced analytics and improved sensor / telemetry will, hopefully, mean even better, faster, cheaper way of producing products.

It is anticipated that the commercial rollout of 5G will further fuel adoption of IoT and related technologies as connectivity may finally become nearly ubiquitous

Sam Cece co-founder and ceo of Swift Sensors spoke to FutureIoT on the implications of technologies like cloud computing on the manufacturing process, including the use of evolving technologies like IoT and NB-IoT.

Sensors, including telemetry, have been around for some time. What benefits/improvement do the cloud offer that is significantly better than legacy implementations?

Sam Cece: The cloud offers several benefits to measurement and monitoring systems.  I’ll address each of the areas separately.

1. Better Reliability -- Clouds services, such as Amazon Web Services, have “five nines” availability and redundancy so there’s never a concern regarding access or storage of data.
2. Better Security -- Cloud services incorporate the leading edge security techniques not available or not affordable to vendors of most on-premise storage systems.
3. Lower Total Cost of Ownership (TCO) -- With a cloud-based system there’s no installation, maintenance, or upgrade costs. The system is always up to date and always has the latest features available without a costly upgrade process.
4. Easier integration with 3rd-party apps, tools, and libraries -- A large and growing community of cloud tools exists for integration of capabilities such as artificial intelligence (AI), data analytics, mapping, and predictive maintenance. Most if not all vendors of these tools offer standard APIs, which allow easy and seamless integration.

Part of the allure of open systems today is the ability to bring in best of breed solutions. In the case of the sensors as used in, for example, a manufacturing environment, is it better to go all out with multi-vendor solutions? Are closed systems still viable in the era of cloud?

Sam Cece: In the era of the cloud, it’s difficult to find or imagine a completely closed system.   Even the large, highly proprietary systems offer some level of interoperability, perhaps through a standard sensor interface (e.g 4-20mA, HART), through a wireless protocol (e.g. BLE, ZigBee, WiFi, or LoRaWAN), or with Restful APIs that provide access to the system’s data on the cloud.

There’s no perfect solution along the continuum of open multi-vendor to proprietary closed systems.  Most important is a thorough level of integration and system testing of the solution. The fundamental elements of the system must work reliably, all of the time.  This is particularly important with Industrial IoT systems. Industrial and manufacturing plants cannot afford downtime and maintenance of an IoT system that is not fully integrated.  It’s the basic starting point from which an industrial customer views an acceptable solution.

We designed our system at Swift Sensors to be fully end-to-end ready to deploy and run from sensor to cloud without any modifications.   We see this is critical to meet our industrial customer’s needs. We build our system on standard technologies, including WiFi, ethernet, BLE, ZigBee, and Web APIs so our system can easily incorporate new hardware and software functionality as well as export data from our hardware and cloud-based software.

A key concern around IoT is the lack of attention paid to securing these devices. How are vendors like Swift Sensor addressing this concern?

Sam Cece: Security has to be designed into the system from the beginning.  The most critical elements of the system are the communication between the wireless access points and the internet.  It is at this interface where attacks and intrusions typically occur. We use SSL encryption in the communication between our Bridge (wireless access point) and the Cloud (internet).  We are fortunate to have one of the leading experts in cybersecurity for banking systems who has architected our entire security system.

There is never a perfect solution and blockade from cyber-attacks, but a continued vigilant approach in the design, and continuous testing of an IoT system can minimize the cybersecurity risk.

Swift Sensors claims that its cloud wireless sensor system that can be configured and deployed at one-tenth of the cost of traditional systems. How do you keep this cost down?

Sam Cece: The significantly lower costs of our system is a factor of three elements:

1. Cloud-based architecture -- Storing data and configuring the wireless system on the cloud completely obviates the hassle of installing and maintaining software on-premise. This also removes the need to purchase and maintain capital equipment in the form of servers, workstations, or on-premise data acquisition systems. Software maintenance and feature enhancements are seamless and silent and do not require any resource.   The total cost of ownership of a cloud-based monitoring system is at least 1/10th of an on-premise alternative.
2. Lower installation and cabling costs -- Wireless sensors, by definition, do not need cabling back to the central data acquisition system. The sensors can be placed in locations not accessible by wired sensors and are not restricted by a cable that could make the installation more difficult.  Cabling costs are eliminated and installation costs are lowered with a wireless sensor system.
3. Lower cost hardware design using high volume wireless SoCs and sensor ICs -- We use a common radio architecture in our sensor and bridge hardware using industry standard ICs and SoCs. The high volume use of these chips reduce the cost of our hardware platform and ensure compatibility with current and future standards, such as BLE, ZigBee, WiFI, LoraWAN, and NB-IoT.

In your view, what will be the key trend for manufacturers to pay attention to in 2019?

Sam Cece: The key question for all manufacturers in 2019 is “How will you implement Industry 4.0 in your factory?”  Industry 4.0 includes improvements in processes including Interconnection--the industrial IoT), Information Transparency--using data to make decisions, Technical assistance--using data and machines to assist humans in difficult decisions and tasks, and Decentralization--autonomous decisions and tasks.

Implementing Industry 4.0 doesn’t have to mean disrupting the entire manufacturing process or paying consultants to implement an overarching (and probably unrealistic) digital strategy.  Our wireless system, for example, allows a manufacturing company to start small, monitoring equipment and process to gain insight into how to improve operational efficiency and a machine or shift by shift level.  Getting started with Industry 4.0 is often the hardest part, and critical for all manufacturers to say competitive.