June 3, 2015
Last Wednesday nearly two hundred people from the MA region came together to share insights on the need to rethink strategies, operations, processes, and technologies when embarking on the internet of things (IoT).
Delivered by John Clippinger, Research Scientist at the MIT Media Lab Human Dynamics Group and Executive Director and CEO of ID3, the keynote address highlighted the transformative way in which devices have taken the place of people as primary observers and data gatherers.
Opportunistically, this enables multiple sources to collect and analyze data providing user experiences and which can have a hugely positive effect on organizations’ bottom lines. And as machines get smarter, the experiences continue to get better.
Yet while the technological capabilities continue to grow, so does the need for defining data ownership, governance, and authentication measures. And understanding that the antiquated regulations and processes must be updated to address our current and future needs.
As Clippinger stated “we are creating new classes of assets” and those all need to be controlled. From the smart televisions in your home collecting data about your TV habits to your cars collecting data about your driving behaviors. This data – your data – is owned by the provider. And you, the consumer, must balance these improved experiences with your sharing your personal data.
Clippinger wrapped his talk with a set of autonomous governance technologies that must be considered when multiple devices are collecting data on multiple stakeholders, across many borders, and a number of owners, including:
- Governance without Government
- Authorization without Authorities
- Regulation with Regulators
- Adjudication and Sanction without Lawyers and Judges
- Polis without Politicians
- Auditing without Auditors
|John Clippinger Presentation|
Following the keynote, we had a number of sessions that took deeper dives into different aspects of business strategies, regulatory bodies, data infrastructure and architects, and security.
In a panel that included Michael Munsey of Dassault Systemes, Andy Thurai of IBM, Rob Purser of Mathworks, Howard Heppelmann of PTC, and led by Chris Rezendes of INEX Advisors provided their experiences in redefining how strategic processes are created and implemented.
Asked how each defined IoT in their own views the panelists talked about better products and experiences, enjoying the benefit of having better insights and interactions into the world that never existed before, and understanding that the value derived is for themselves, their customers, and ultimately their customers’ customers.
Heppelmann expressed how IoT has changed the shape of business in that the conversations are happening at the CEO level as they flow back to engineering, sales and marketing, and every aspect and department within a company. Munsey followed that up with the huge capital investment that is going to be required, and the multitude of infrastructure providers that are gambling their futures on supporting IoT infrastructures.
Similarly to a data science team, embarking on IoT will require a variety of people across the organization each with different skill sets to build and manage a connected product. Thurai furthered this point by stating that often it is the CMO or others who want to drive business value and revenue funding these products.
|Dassault Systemes Case Example|
|Mathworks Case Example|
The next panel, What Big Data Will Require in an IoT World, moderated by Paul Barth from Podium Data, included Kris Alexander from Akamai, Chris Baker from Dyn, and Pavandeep Kalra from Microsoft.
Alexander talked about how the entire model of big data has been inverted with IoT, whereas the internet has been geared to send massive amounts of large data sets out, but connected devices are now sending massive amounts of small data sets back in.
Kalra who focuses much of his work on machine learning, talked about how more and more there will be a marketplace of machine learning APIs which you can leverage. He also talked about pushing everything to the cloud and then testing in that type of safer environment before making any operational changes.
Baker talked about the importance of obtaining the right information and putting that back into the device to continue to learn and redeploy. Baker also said that having a firm understanding of timing when collecting and redeploying data, such as a pace maker, is the most important component of an IoT product.
Our next session was a fireside chat style discussion between Niko Pipaloff of PwC and Said Tabet of EMC, who was there on behalf of the Industrial Internet Consortium (IIC).
Tabet talked about the IIC’s mission, which is to take IoT forward. Just a year old, the IIC has already amassed 170 member companies and is currently running 6 test beds across the globe to help develop the use cases that companies can use as an example of how to bring the physical and digital worlds together. . The IIC’s test beds can leverage use cases to build the requirements for the future.
Pipaloff spoke of the benefit of having already proven that working together and offering open standards at with the creation of internet over two decades before will help us.
They concluded the discussion by reiterating that IoT is massive and it’s here already and that a number of small organizations must all come and work together to create the standards and regulations
As the Value of Data panel got underway, it was very clear that it really is all about the data. But what this stellar group talked about was how to understand what data you need and why.
According to Rob Patterson from ColdLight, 9 out of 10 companies coming out of academia are based on artificial intelligence technology - highly data intensive. Combining the AI with machine learning and skilled talent, the intelligence gained can be enormous.
Don Schuerman from Pegasystems, talked about the best insights coming from multiple devices connected to each other to discover external drivers and changing the outcomes.
Subu Ramasamy from Philips Lighting talked through a number of case studies in which creating connected lighting could provide efficiencies through measuring occupancy rates, not just in energy consumption, but also in cleaning or stocking particular offices based on use. He went on to say, customers are not looking for technologies they are looking for services.
When probed by the moderator, Jeff Kaplan of THINKstrategies, the panel members all agreed challenges were around the complexities surrounding the variety of data sources and sets as well as the the limited talent that exists today.
The conference closed on perhaps the most important topic of the day, security. Editor in Chief of Security Ledger, Paul Roberts facilitated a discussion between Brandon Creighton of Veracode and Paddy Srinivasan of Xively by LogMeIn.
The panelists talked about the first step – identifying there is a very real problem. And they stressed that today we are in a world where every vertical now needs to have an information security mindset, but unfortunately that is not the case.
Srinivasan and Creighton spend a great deal of time talking with their customers about what risks exist in products that do not have operating systems.
Srinivasan went on to suggest that he urges customers to follow a multi-step process that includes: uniquely identify each connected device, then securing the information that is coming from the device to the cloud. And once the device and communication are secured, there is then a need to secure and ID the data and assign permissions and actors to that data.
Creighton concludes his remarks with stressing that until there are established protocols for IoT, to have have your engineers do security training, understand what the basic code mistakes are, and then retain security experts to review the architecture and communications networks. And the most important thing is to do this during the development, not as an add-on.