#IoT panel: software is king and you need to understand the industry when building applications
Posted on October 14th, 2015
10/13/2015 @Verizon, 140 West Street, NY
Samantha King moderated a panel of experts on the #InternetOfThings. The discussion covered many aspects of mobile computing which can be divided into the 4 areas of smart: wearable, home, city, enterprise.
The panelists came from different parts of the IoT world.
Jamyn Edis @Dash talked about how Dash developed onboard diagnostic device that plugs into the car dashboard and broadcasts the performance of the car and driver to a smart phone. They have a deal with DOT in NY and an insurance company. The device also allows fleet managers to monitor driving behavior. They collect sensitive data, but they don’t share the data with others.
He talked about the importance of a software platform to capture the data and how they want to be hardware agnostic.
Ed Maguire is a research analyst for CLSA. He divides the industry into consumer and industry sectors. He believes that sensors, wireless connectivity, data connectivity will soon be absorbed into household products. But, sensors in industrial processes might have an even larger long term effect. These sensors can help us better understand what is going on in complicated systems and help us improve the processes.
He sees a glut of smart wearables. The next area will be the smart home: Savant announced a $500 universal remote. GE, IBM, CSCO are pointing their businesses toward IoT.
He cautioned that when industrial and public systems are linked and controlled by sensors there is risk and we need to be careful about what is revealed about this infrastructure.
Ed said that devices should run off a platform and solve a specific problem. This means that specific industry knowledge is needed before developing the device. Industry knowledge is also important in considering the adoption of a new method. For instance, connected farming offers many advantages but might be a difficult sale since farmers buy through channels and they are not comfortable when someone else owns their data.
One trend he sees is edge analytics where as much processing/analysis is done near the source so the data send over the internet is minimized.
Ted Ullrich consults with hardware startups. He sees a proliferation of sensors connected with wifi and cellular. Examples include Hitatchi instrumenting its trains, Canary developing smart apartment security system, Kean Home working on connected ventilation and sump pumps.
Ted sees some companies using open source and others using closed source depending on their changing needs. He mentioned that originally Makerbot was open source, but then became closed source as they wanted to control their hardware.
The largest changes will be in software and connectivity as sensor technology evolves more slowly.
Thomas Gilley is currently in his 12th startup. Software is the key to useful outputs. In the next 3 years he sees large companies waking up to what is happening in small companies and acquiring them.
Data is the thing of value. For instance, a soft drink company has connected their dispensers and are looking to correlate demand to events. They would eventually make predictive models of demand.
To protect your ideas, start a dialogue with larger companies and get them on board. Once you become larger, create an advantage in execution.
Real time stream processing is becoming increasing important. Data cleaning is an essential part of this processing.
You need to understand the domain to know what the sensors tell you.