Did you know that honest brokers persist in technical domains? The thing is; “reality” forces agreement. Bridges either stand or fall, pipes leak, colors run. Physical systems punish ideological games. You can still make mistakes as a buyer, but you can mitigate them by being prudent. The strongest IoT ecosystems operate like infrastructure, not software products.
In IoT an honest broker is not a company or a piece of software, it’s an architecture. It is a systemic element. A company is focused on profitability. The way to make a profit is to keep your customer around, sometimes blind, gagged and bound.
Your system design should allow devices from multiple vendors. It permits data portability. There might be a focus on replaceable components. Communication protocols are secure, fit for purpose, the functionalities should not be opaque. You are able to use your data, to work with it. Another protection for data streams is leaning on standards that have a wide user base and good support; MQTT, LoRaWAN, Modbus, etc.
Practical steps: Own the data pipeline
Your data should flow to your environment. Now, we can get into technical discussions about scope: all of it, some, or just reports? In any event, you should be able to access it. A strategically dangerous concept is “device – vendor cloud – vendor dashboard.” At both ends of the data pipeline, you’re locked out.
Many vendors want to sell you a bundle. That is not in itself a bad thing. At the beginning this is handy. You grow up inside the system. You get sensors, gateways, connectivity, dashboards and analytics that will work together. Think a little bit laterally here; can each layer be swapped out and the systems still remain flexible and functional?
Gateways are particularly important in this regard. Gateways are gatekeepers. A good one prevents one device manufacturer from becoming a dictator. Flexibility should allow different devices, protocols and networks while retaining secure communications. Dashboards are the other end of the data pipeline, but the gatekeeping element is similar. A good dashboard gives clean data access, the ability to filter streams and port out data sets to back end tools.
Human in the loop
A lot of issues with tech systems center around human organization, not technical limits. Fear of complexity is an obvious one, but so too is limited training.
When your procurement team goes for one vendor to simplify contracts, it automatically creates a power dynamic on the technical side. Interoperable frameworks do not hold much water on a spreadsheet, but these are real-world value propositions. Make processes auditable, predictable and reversible.
If we think in terms of slightly longer sight lines; we are looking at surviving vendor changes. Can we maintain data after vendor company bankruptcy, takeover or merger? Can we move through different hardware refresh cycles? Do we avoid subscription traps? Does our system architecture keep us aligned with reality, as a compass does on a mountain afternoon when the clouds come rolling in?
