ERP News

IOT

Eight IoT tips to ensure project success

659 0
View the original post on site

Eight IoT tips to ensure project success

A rundown on how to get started with the Internet of Things

 

The Internet of Things (IoT) is one of the most talked about trends in the IT industry. It is touted as a way to make companies more efficient and streamlined, and is at the centre of big data use.

But getting involved in the IoT can be daunting as there are major hurdles to overcome and no clear way to get started.

To help solve these problems, V3 has gathered eight tips from analysts and industry experts that should guide forward-thinking IT and digital leaders in embracing the IoT with success.

8. Join the IT dots
Ocado smartbots will pick items from thousands of crates

Building an IoT environment stuffed full of sensors is one thing. Ensuring that they can all be controlled easily by a central system is another.

Tim Ensor, head of connected devices at Cambridge Consultants, explained that this is often overlooked.

“It’s assumed to be easy because we’re all so used to smartphones, WiFi, Bluetooth and so on, but in a lot of applications that we get involved with it’s far more challenging than people realise,” he told V3.

Ensor advised businesses to acknowledge this difficulty and set up a robust central control system from the outset.

An example of this can be seen in Cambridge Consultants’ recent work with Ocado to create a wireless network in a new fulfilment centre to allow 1,000 robots to ‘talk’ to each other and a central system as they move around the environment (above).

There are also ‘out-of-the-box’ products, such as Microsoft’s cloud-powered Azure IoT Hub, that act as middleware to connect IoT devices to back-end systems.

7. Invest in streaming data
Big data

Managing the sheer volume of data produced from IoT networks and devices can be an IT nightmare, leading to huge amounts of information that needs lengthy processing and analysis before it can be put to use.

Craig Wentworth, principal analyst at MWD Advisors, suggested that companies should invest in streaming analytics systems that crunch data as it comes in rather than simply push it to a large database.

“You need to invest in tools that handle streaming data if you’re planning to act in the moment and respond to what your IoT devices are telling you pretty much as it happens, because the safety or operational efficiency of your machinery depends on it,” he told V3.

“That’s less important if you’re just looking to hoover up as much data as you can and then pore over it for retrospective reporting with some time lag. But the chances are there’s more value to be had by acting on insight quickly.”

There is plenty of steaming analytics software available for companies looking to invest in this area, such as the Vertica big data analytics platform and a similar IoT-oriented system from Splunk.

6. Embrace embedded intelligence and automation
ai-robots-waters0316

Collecting, processing and analysing IoT data can be an arduous task just to get to the point where actions can be take on the information. Automation is the key to pushing past this barrier.

Machine learning analytics tools, such as the cognitive computing capabilities on offer from IBM’s Watson, can automate the collection and analysis of data.

In some cases this can enable machines to act on analysed data autonomously, such as triggering a warning when a machine needs maintenance before it malfunctions.

Ton Steenman, vice president of Intel’s IoT division, suggested that IoT networks should be embedded with smart analytics capabilities at device level to enable this automation.

“It is necessary in industrial applications to have what we call intelligence at the edge, because you cannot always rely on the network connection being there,” he said.

“You need a certain amount of intelligence to guarantee behaviour, particularly for mission-critical applications like smart grid or water supply management, or transport [services] where you need a real-time response.”

5. Don’t obsess over all the data
A lady in front of a blackborad with lots of squiggles denoting stress

It can be tempting to analyse all of the data collected from IoT networks to see what results it may yield.

But MWD Advisors’ Wentworth said companies should not try to evaluate all data, and should instead focus in on the aspects of it that have an intrinsic value to business operations.

“Don’t obsess about knowing the measurement of everything but the value of nothing. You can pretty much assume that there is, or soon will be, a sensor out there which can get you data on whatever aspect of use you’re looking to capture. These days, that’s the easy bit,” he said.

“And if you can process and store what you need effectively, the raw technology challenges are dealt with.

“But what that leaves is arguably the more important bit about modelling what you’re being told about and applying analytics in the context of your particular business problem so that it can tell you what you need to know.

“If you’re not careful you can get too hung up on the number of sensor points you’re hooked up to, or the rate you’re sampling at, when what you really need to anchor your IoT initiative is a really good handle on your business issues and how data, and what data, can add value. Don’t let the data on its own be the tail that wags the dog.”

4. Seek out the right skills
big-data-skills

Having the hardware and software to set up an IoT network is one thing. Having the right skills to get the most out of it is another.

Cambridge Consultants’ Ensor advised firms embarking on IoT projects to appreciate that it is a multi-disciplinary challenge and needs different skills from different departments.

“If you end up with just the tech discipline trying to run [the project], for example, you end up with bias. So you have to get everyone in a room and work together to manage the requirements,” he said.

Jim Tully, an IoT specialist at Gartner, echoed this advice. “Recommendations we tend to make are that it’s important to put together a multi-disciplinary team made up of different people from different departments,” he said.

“If it’s a manufacturing company you want someone from manufacturing, someone from resources, someone from accounting. Not a big team, maybe six people, who are then tasked with identifying how the organisation can use the IoT.”

3. Secure it
cybersecurity

Security is vital when embracing the IoT much like the deployment of any new IT system. The number of end points being opened gives hackers more attack vectors to infect with malware or use as a backdoor into a company’s network.

The FBI has suggested that IT teams should have a clear understanding of the type of password and protection that come as standard with IoT devices, and work on boosting security and ensuring that access to WiFi networks is tightly controlled.

“Many routers give you the option to set up more than one network. If yours does, separate your computing devices from your IoT devices and spread them throughout several different networks. That way, if cyber criminals break into one network, the damage they do will be limited to the devices on that one network,” said a post on the FBI blog.

The agency also advised companies to be careful when buying IoT hardware. “Purchase IoT devices from manufacturers with a track record of providing secure devices, and set your devices for automatic updates when available,” the post said.

There are several security products that can protect individual devices through to entire networks. Webroot’s IoT Security Toolkit, for example, uses cloud-powered intelligence to detect threats across an IoT ecosystem.

2. Embrace the cloud
shutterstock-91647578

Cloud and the IoT are good bedfellows. The data that can be collected and processed from an IoT network is best partnered with cloud-based storage and software that can scale up or down to cope with different levels of demand. This, in turn, helps to manage compute resources and running costs.

MWD Advisors’ Wentworth explained that cloud comes into its own with the IoT when companies are looking to carry out fast analysis on large amounts of data.

“If you’re sampling some aspect of device behaviour at high frequency, because that’s what you need to give you the granularity you’re looking for, you’d better have enough storage and processing power to accommodate extreme bursts, and act on what the analysis can tell you fast if you want it to make a difference. That might mean your [IoT] big data strategy needs to embrace cloud services, where you dial up and down capacity as and when you need it,” he said.

The proliferation of cloud services, and continuously dropping prices from providers such as Microsoft, Amazon and Google, mean that enterprises have several cloud options and can choose the most suitable for them.

1. Identify the business angle and strategy
shutterstock-341846936

Deploying an IoT system is pointless unless it provides clear benefits for the business, such as reducing unexpected maintenance on assembly lines or improving the speed of product distribution.

Gartner’s Tully advised companies looking to get into the IoT to learn from others that have already deployed an effective network.

“[Companies should] look at what other organisations are doing. They should look at totally different industries to get ideas from them as well to eventually form a view of the different ways in which the IoT could be used,” he said.

“The IoT tends to be used for two fundamentally different purposes. One is for internal streamlining, efficiency and cost-savings, and the other is for external-facing, customer-facing, revenue-generating purposes.”

A clear strategy focused on business outcomes over technology use will allow companies to achieve a better return on investment and avoid the pitfalls of creating nothing more than a vanity project.

Leave A Reply

Your email address will not be published.

*

code