Picture credit: SAP

Opinion Industry leaders in each sector are carving out their share of the IoT market. The latest to stake a claim is SAP, the world’s largest inter-enterprise software company and the world’s fourth-largest independent software supplier, overall.

When the company was founded by five former IBM employees in 1972, the original premise was to provide customers with a way of interacting with common corporate databases. Now, the most common application of SAP software is to run internal business operations; both IBM and Microsoft use SAP applications to run their enterprises.

With an established claim on business-level software, it is easy to see why SAP would be working feverishly to ensure that they do not lose ground as demand for IoT business solutions increases. SAP recently acquired PLAT.ONE, an IoT platform/solutions provider and Fedem Technology, an analytics software company. Both companies are being integrated into SAP HANA in support of their launch of SAP IoT, set to focus on “applying machine learning/advanced analytics to the vast amount of data that IoT devices collect.”

SAP IoT in action

The company is already testing the waters in a few applications. A mining company in Russia uses SAP IoT to monitor the health of mine workers in an effort to reduce safety and health risks on the jobsite. Employees undergo a health screening, performed by a robotic device, prior to each shift. The results of the screening are fed to mining leaders and used to calculate potential health and safety hazards and the long-term impacts of environmental working conditions on employees.

In Japan, a public transit company is using a connected sensor in combination with weather monitoring, traffic monitoring and other data to increase the safety of their commuters. Information on driver behavior and biofeedback is delivered to a monitoring center where alerts are created when a potentially unsafe condition arises. This information allows the monitoring facility to respond with plans that promote the safety and well-being of the passengers and drivers.

SAP is even working in the energy sector in Norway where IoT devices are connected to wind turbines in the field to feed data back to engineers. The teams use the data to power scale-models of wind turbines and analyze the potential impact of weather conditions and changes to the design using real-world data. Meanwhile other industry giants are staking an IoT claim.

General Electric supports industry

In the industrial space, GE has become a clear frontrunner. The company’s initiative, dubbed “Power of the 1%” is predicted to save billions of dollars over the next fifteen years; GE claims that a 1% increase in efficiency will save the oil and gas industry $90 billion; aviation $66 billion, healthcare $63 billion and rail $27 billion. These efficiencies are driven primarily by GE’s Predix platform, a PaaS solution that is billed as laying “the foundation for the world’s first and largest marketplace for industrial applications.” In its mature state, Predix will bring together industrial data from multiple companies and applications to drive better understanding of field data, improve designs and decrease the financial burden associated with maintenance.

Google covers the home

Google has already been incorporated into the daily lives of many users. We use it communicate in real-time with friends and family via Hangouts, Google’s Calendar integrates with nearly every other calendar platform making it easy to plan family events and keep tabs on schedules; and through the purchase of Nest, Google is building a connected home experience that will be driven from our Android phones.

IBM is building the backend

Supporting much of the connected technology is IBM’s Bluemix portfolio; a service that gives developers the tools they need to “quickly and easily extend Internet-connected devices to the cloud to not only leverage data from, but also to build an IoT application in just a few minutes.”

This capability is giving companies the power they need to adapt existing technologies to compete in a cloud-based world.

While SAP may be the current front-runner in connected business solutions, history has shown us that it does not take much to lose footing in this competitive space. The major players in each sector primarily gain ground through the acquisition of companies that have perfected specific solutions in their niche and then incorporating the intellectual property into their own products. Maintaining leadership status requires a combination of development activities and intelligent acquisition of best-in-class solutions.

(c)iStock.com/Varijanta

The introduction of connected machines into industrial environments has raised quality standards, led to increased profits and improved the maintainability of both manufacturing equipment and end products. Manufacturers that have integrated their production floors with other aspects of the business (including design, sales and supply chain) are seeing the largest benefits as the machine learning aspect of connected networks trickles into all areas of the business.

Here is a look at six ways machine learning is impacting industrial business.

Changing the face of customer relationships

One clear indicator that machine learning and artificial intelligence are coming together to improve customer relationship management is Salesforce’s acquisition of several Machine Learning and AI companies. Since 2014, Salesforce has acquired six AI and Machine Learning companies including: RelateIQ, TempoAI, MinHash, PredicitonIO, MetaMind and Implisit Insights. As a result of these acquisitions, Salesforce has released several new products leading to an estimated new product revenue of $635 million by FY18.

Dramatically improving both product and service quality

Product quality and customer service are woven throughout every aspect of a workflow cycle. Production cell leaders impact customer service by ensuring that products move smoothly through their cell and that waste is minimized, thereby reducing costs. Sales team leaders ensure product quality by understanding customer needs and working with design teams to develop best-fit solutions. With machine learning, executive teams are gaining a better understanding of how decisions both upstream and downstream of specific points in the production cycle are impacting product and service quality.

Optimising processes with greater accuracy and better results

The fast-paced world of manufacturing requires leaders to constantly consider the impact of each decision and to make trade-offs based on schedule demands, material and machine availability and customer needs. Prioritizing each demand while simultaneously managing waste, equipment efficiencies and human resource efficiencies has always been a challenge to manufacturing floor leaders; optimizing each of these aspects to improve yields and profits is a careful balancing act. Quick access to reliable data dramatically improves the ability of leaders to make the best decisions.

Improving price competitiveness without compromising profits

With so many manufacturers available, the demand to provide high-quality products at the best possible price has never been higher. Integrated supply chains, especially those that have connected some aspects of their own internal systems with those of their vendors can provide customers with variable pricing that closes the deal while maintaining the margins the business needs.

MaaS (manufacturing as a service) and on-demand manufacturing are becoming a reality

As individual departments within the business are integrated, the next logical gaps to close are those that exist between the end customer, the OEM (original equipment manufacturer) and material suppliers. The benefits of subscription services (consistent pricing, reliable service, scalability) are trickling throughout all aspects of commercial enterprise. End-customer orders are driving demand while data collection and machine learning are making it easier to anticipate these needs. Because of this data, production runs, even those of highly-customised products, are quickly scalable.

Improving predictive maintenance analysis and driving efficiency through maintenance, repair and overhaul (MRO) stations

Maintenance has always been a cost-driver in industries that rely on expensive equipment in the field (think airlines and major shippers). In these industries, performing field maintenance at just the right time (before the equipment fails but not so early as to scrap excess product life) dramatically impacts profitability and safety. Equipment that must be returned to an MRO station is under even greater scrutiny as this may mean that part of the fleet is inaccessible until the equipment is returned. Machine learning and data gathering are dramatically improving maintenance scheduling, reducing equipment downtime and driving greater profitability.

The Industrial Internet of Things is about more than simply connecting the machines of one area of the business and even about more than connecting various departments – IIoT is bringing the benefits of iterative algorithms to the business as a whole. Data is gathered, analysed and used to make minor changes at specific points within the lifecycle of the product. Those changes are then either accepted or scrapped based on the results of additional analysis. Just as software testing involves many iterations, industrial manufacturers are finding success through the continuous improvement machine learning enables.

(c)iStock.com/mattjeacock

The growth of all things connected and of all things cloud seems unstoppable, and monikers are jumping to the logical end…that the Internet of Things will be the Internet of Everything and that the ‘as a service’ model will morph into ‘everything as a service’.

On-demand products have always had a place in business. For more than forty years, mainframe time has been sold ‘as a service’ where users only pay for what they need when they need it. At a time when businesses are growth hacking their way to larger and larger client bases and new companies are coming online every day, the on-demand model provides the perfect balance of functionality and value.

XaaS is defined – by TechTarget – as “a collective term said to stand for a number of things including “X as a service,” “anything as a service” or “everything as a service.”” XaaS could just as easily refer to CaaS (communication as a service) as it could PaaS (platform as a service). As long as the service is delivered over the internet and not installed locally, it meets the “aaS” requirements.

Service game

The “aaS” list just keeps growing:

• Infrastructure-as-a-Service (IaaS): hardware, software, servers, storage and all of the associated maintenance is handled by someone else. Generally, this means that many of the tasks are automated by the provider and that both monitoring and systems management will require more time.
• Desktop as a service (DaaS): the infrastructure of VDI (virtual desktop infrastructure) is hosted in the cloud. It is a lower cost alternative to VDI yet offers many of the same features: data storage, backup, security and upgrades.
• Disaster recovery as a service (DRaaS): While data is sent via the cloud, it does not necessarily remain stored there in DRaaS solutions. The provider may offer cloud-storage or the data may be sent to a server to lie in wait until disaster strikes. Businesses that utilise DaaS often rely upon it to function as a disaster recovery plan instead of enlisting a separate service.

Benefits of ‘as a service’ models

What is it about the “aaS” model that has businesses trading assets for memberships? In a word: flexibility. Businesses are tired of carrying technical debt and instead want their technology choices to remain as agile as the rest of their practices. The lure of a plug-and-play solution is that it eliminates the need for in-house support personnel while ensuring that tools and support are available on a moment’s notice.

Perhaps even more attractive is the flexibility to choose the right solution provider at the right moment. No longer will large monopolies of “compatible” or “preferred partners” rule decision-making. Instead, businesses have the freedom to choose the provider with the best set of features for the best price.

The future…as a service

Technology startups of today will never have known the limitations of release specific software. The frustration of buying a software suite only to have a new release hit the shelves days after a major investment is made. While these businesses are young by today’s standards, according to HfS, “the make-up of the Fortune 500 in five years’ time will be very different from todays.

One of the key reasons for this is the rapid evolution of new business [is that] all their services are in the cloud and their entire infrastructure is delivered to them on a seamless “as-a-Service” model.” HfS founder, Phil Fersht goes on to write that these businesses will see significant cost and speed-to-market advantages thanks to the ‘as a service’ model.

(c)iStock.com/chinaface

By Alex Vilner, managing director, SaM Solutions

With all things tech switching to cloud-based platforms the ‘as a service’ industry is exploding. The one that started it all, SaaS (software as a service), is now joined by IaaS and PaaS.

The commonality for these ‘as a service’ offerings is that they are all based in the cloud, solely managed by the provider and available to clients via a membership platform.

Connected devices are changing the landscape of technology. We don’t purchase anything – we subscribe to it. The pace of technology growth led to a culture of obsolescence. Barely had companies adopted and incorporated a new piece of software or infrastructure into their technology stack and a new version was released.

Membership-based or ‘as a service’ offerings do two things: first, they ensure that companies always have access to the latest features and services without the cost and headaches associated with incorporating new releases. Second, there is an economy-of-scale dynamic at work.

Developers of these services work on the offering every day. Building a feature or improving an existing one might take two or three dedicated developers working for weeks. The ROI for this investment, if the services are maintained in-house, simply isn’t there for most companies; however, service providers push this new feature to all of their customers meaning that the benefits are experienced across the board.

IoT reliance on the cloud and the benefits of the ‘as a service’ model has made companies re-examine their available tools in search of solutions that are as flexible, scalable and economical as their own products.

PaaS operates in the area between SaaS and IaaS; giving full control of the application and the data to the developer. The infrastructure (middleware, operating systems, virtualisation, storage and networking) are unlikely to change during the development process of an IoT application. This is why it makes sense to leave those aspects to be managed by someone else (IaaS provider).

Why does PaaS fit with IoT?

Data, data, data: So much of the IoT is about data. Data collection. Data storage. Data analysis. Much of the testing required by IoT applications requires different configurations of data and most IoT applications are developed for multiple use case scenarios. PaaS allows developers complete control over collected data without the burden of managing storage systems.

Future flexibility: Data analysis requirements will change as the application sees wider adoption and developers learn more about how users are deploying the product in real life. Smart developers plan for this eventuality by building products that include flexible components. PaaS allows developers to customise data analysis both now and in the future.

Pace and competition: With predicted billions up for grabs, rest-assured that competition in the IoT space is only going to continue to grow. Releasing products months and sometimes years ahead of pre-IoT project estimates requires development teams to reconsider which aspects of the stack need to be built in-house and which can be managed by someone else.

Workflows: For many adopters, the primary draw to IoT solutions is automation. Instead of relying on the inspector to manually bring up the inspection requirements for a component, a device scans a barcode on the part and ensures that all of the requirements and data recording areas are automatically displayed to the inspector when the component arrives. Developing new workflows and optimising existing ones is a driving force behind new releases of IoT applications. PaaS provides just the access needed to continue making these improvements.

Virtual everything: Not only have our tools gone to the cloud, our teams have too. Most development teams include members from various facility locations, both on- and off-site, and help from outsource firms. Teams are rarely co-located making tools that are based in the cloud necessary to efficient team operations.

According to Fabrizio Biscotti, research director at Gartner, “the PaaS segment showed impressive growth, not just in the AIM (application infrastructure and middleware) market but across the entire enterprise software market” in 2015. This growth points to the increased demand for IoT-supporting development tools and the continued growth expected in this area.