IoT is Here to Stay
Crowe recently completed an IoT and ERP project with Motis that provides the ability to read data from their machines receive alerts and maintenance updates and share some specific machine data with others in the cloud. Our solution involved adding sensors to a few “dumb machines” and connecting them to the Azure IoT Hub via Raspberry Pi and Arduino routers, as the video shows below:
With this project's success, we wanted to share items of note when thinking of IoT for your own business and ERP platform:
IoT Considerations:
IoT is a massive buzzword
So much data --what’s critical is key
Customers and vendors all agree that IoT is on top of everyone’s mind and on many business agendas. Most of the IoT Technology and Solutions are in the “Peak of inflated Expectations” on the Gartner Hype Cycle, while some really important related technologies are past that point and truly are in use. Many forward-thinking organizations are adopting it, giving them an advantage in the next 2-5 years in their industry. They will be at the forefront of disruption of some old industry models by investing and learning now. Therefore do not toss this aside as some fad; leaders should take note and take careful steps in embracing IoT within their ERP systems.
Be Aware of the Cost
Cost per machine is not necessarily the concern – its multiple machines that multiply the costs.
Although many sensors are relatively cheap and easy to hook up, think about the actual cost of IoT from the beginning of the project through to the end. IoT is hard, dirty work, and getting data from old machines to your ERP is not easy. The physical installation of the sensors can be costly and time consuming from design, to setup, testing and final deployment. Some of it is repeatable, but many machines communicate differently which requires tweaks to software and data flow. Multiply this over the number of machines in a business, and the numbers add up. Also, be sure to keep equipment manufacturers aware of what you are doing, as it is possible to void a machine’s warranty because of the modifications.
Leadership and Skills
New skills and positive leadership required
Others before have stated it -- the digital age requires different skillsets and leadership. About 80-90% of IoT can be covered with simple alarming and tracking while the other 10-20% requires the attention and involvement of a true Data Scientist -- often people with a PhD or Master’s degree who specialize in Machine Learning. This is not something you learn on the fly. The algorithms, frameworks and dataflow involved in some of these solutions require a new level of skills and knowledge. Organizations like Crowe are investing heavily in hiring and training the right people in order to have the right skillset to build effective solutions for our clients.
Which brings me to leadership. While we are currently in “Industry4.0”, we are moving rapidly toward “Industry5.0” where we no longer just connect machines with each other but actually using smart technology to connect the human with the machine. This is happening at a pace of change we have not seen before. Leaders need to throw out their old way of thinking and get ahead of the curve, because their competition is all over it. Leaders need to prepare their teams for what is coming, not with fear but with encouragement to identify the opportunities that for many could change or improve their careers. They also need the ability to explain changes to stakeholders and shareholders who might fear having data in The Cloud. Increased interaction between humans and machines in Manufacturing and Distribution, utilizing the human cognitive and critical thinking skills brings change that requires careful guidance.
Keep Pilot Programs Focused
From what we have learned in our projects, it is very important to start out with a few small pilot programs before deploying an IoT project in a business. The pilots should be small and have a very specific focus on two - three KPI’s. Perhaps CDI’s (Critical Data Indicators), is a better term. While technology changes so quickly, it will be easy to fall into the trap of wanting to introduce the next shiny object to the solution, but resist the temptation or you’ll end up in a never ending Pilot Program without any results.
Use the pilot programs to identify more about the business, the technology opportunities and possibilities. Depending on needs and size of the organization, there will always be another Pilot Program to push the organization forward, so it is important to keep the focus narrow for the current project, and deliver small wins. These wins will be the key to long-term success and ongoing support from both the Business owners as well as the people that use the new technology.
There Are No Standards
There is a need to Self-Regulate
In a recent conversation with our Chief Data Officer, I was reminded of the importance of standards around Machine Learning. Standards provide a foundation for faster, repeatable, secure and reliable processes. It further protects our company’s reputation, by using our own approved standards and models. The industry itself has no specified standards resulting in a wide variety of solutions and ways of doing things. It is shocking to learn how many IoT solutions in place today have created a huge security risk through open ports and unsecured communication models. There is clearly a high requirement for self-regulation and standardization, combined with security for the IoT Industry. Therefore, I urge you to check with technology providers and partners about their methodologies and commitment to Security and Data protection.
While there is still a lot of work to do – as is always the case when new technology presents itself -- I am actually very excited about all of this. I see enormous opportunities for our customers across multiple industries, with rapid evolving technologies and solutions. Feel free to reach out with any questions regarding IoT, Support, and ERP - I look forward to it!
Ronald Haantjes
Crowe, LLP
www.crowe.com/supportservices