Chris Murphy from InformationWeek has published a fascinating article about some of the challenges organizations face when trying to implement IoT solutions. The article focuses on Industrial IoT solutions and is based on interviews with GE, Union Pacific and others. I certainly recommend reading the article.
I thought it might be interesting to look at each of the challenges presented in the article to see how the Open IoT community might help solve the problems.
1. The data isn’t good enough.
Getting smarter about how to collect and use the data is still a work in progress. The last quote ‘linkage is one of the next big areas to look at’ is particularly relevant. This is why open standards like MQTT, CoAP and others will be so important for IoT. We need to break-down the existing proprietary protocols so it is easy to flow the data from a Nest thermostat or smart-meter into the demand generation of a power generation plant.
2. Networks aren’t ubiquitous.
“The future Internet of Things model often will combine on-machine processing for urgent needs and batch-data uploads for less timely analysis. Bill Ruh, VP of GE Software, describes this as “real-time, big data processing at the machine. We don’t have anything like that today.”
A lot of the current web and enterprise development architecture assume the existence of a reliable network, in IoT that is not the case. Pushing application logic on-machine or to a gateway will become essential. Cisco has certainly observed this challenge and that is why they talk about ‘fog computing’. It is also why the Eclipse Krikkit project is so important. IoT gateway solutions will also help with this challenge. Eclipse Kura and Mihini provide open source solutions for people wanting to build IoT gateways.
3. Integration is tougher than analysis.
Companies spend 90% of their IoT budgets on those kinds of integrations, leaving insufficient money to drive the operational changes that actually produce the returns, says Ton Steenman, Intel’s IoT business leader.
Building IoT solutions is hard work. It requires the skills of hardware architects, embedded software developers, network engineers, enterprise software developers and data analytic experts. Each solution is often custom crafting of a specific solution. This is where the open IoT community can deliver a huge benefit. We need to collaborate on building a common set of building blocks that people use to build these solutions. We need to use software in help with the complexity of integration. Communities like Alljoyn and Eclipse IoT are building these foundations to help with integration.
4. More sensor innovation needed.
It would seem the open hardware community is well positioned to take up this challenge. We are seeing tremendous innovation in hardware for the maker and IoT community. My guess is that we will see more and more innovation in this community.
It is interesting to see mention the concept of a software-defined sensor. I need to do some more research in this area but like software defined network or software defined radio, my guess is that open source software would be a great way to provide software-defined sensors.
5. Status quo security doesn’t cut it.
“The biggest fallacy is that traditional IT security solves operational technology problems,”
<sigh> Everyone sees security as an issue for IoT but I don’t see anyone stepping up to solve it. </sigh> Someone needs to start working on this or maybe someone can point me to a community working on this, please.
I really like this list of challenges since it is based on companies trying to implement real IoT solutions. There is still a lot of work needed to create technology to solve these solutions but in confirms my belief that the vision we have for Eclipse IoT is focused on the correct customer requirements.