A data-sharing community. It sounds like a simple concept.
But it's not simple if you're dealing across multiple organizations, each prizing its own data. Almost anyone could list the obstacles: security, privacy, identity management, ownership.
Today we introduce the concept of AT&T Network 3.0 Indigo. It's our major push to create a trusted environment where organizations can share data and collaborate on analytics.
Imagine this simple example. A city has an electric utility, an internet company and a major heating/air conditioning repair company. They join a technician dispatch community to share their data - such as vehicle, traffic and appointment data. Through cooperative machine-learning and the broader data set, they get better and better at timing their dispatches. Their work becomes more efficient and customers are happier. And they are still able to keep their proprietary information safe.
Now imagine a data community that would be more complicated but is quite real and exciting: healthcare. Doctors, hospitals, pharmacies and researchers all want better, trusted ways to share data while maintaining patient privacy. This could speed the cure of cancer and other diseases. Patients could participate in medical trials far away from university hospitals.
At the AT&T Developer Summit in Las Vegas today, we announced the Indigo concept. And we invited the audience to think about how they'd like to contribute. Look for a technical summary paper soon, along with a mechanism to share ideas.
Much of the public may think of "the cloud" as a single space. But any business knows better. Data in the cloud is locked in silos intentionally, so it can't be shared outside an organization or between clouds. This stymies the important global trend toward open innovation and open algorithms. Machine learning and artificial intelligence simply work better when they can train themselves on more data.
In other words, we’ll all do better if we can overcome the very real hurdles to sharing.
We see AT&T Network 3.0 Indigo as having three pillars:
- A new platform to enable closed communities of data sharing and collaboration. This platform will have strong identity management, a policy engine (rules enforcer), rightful retention of data ownership, and 2 types of analytics – built-in and bring-your-own. Organizations will create communities and invite others to join.
- Running on a virtualized network, such as the AT&T software-defined network. The AT&T SDN has API access, built-in secure access channels like NetBond, and identity and access-management services. As other networks adopt our open-source ECOMP framework, community memberships can span across networks.
- Necessary infrastructure services – compute, network and storage. Any virtual private cloud provider with API-accessible cloud orchestration could enable these services.
It will all come down to earning trust. Trusted identification, network connection, cloud provider and network operator. We think new advances will help solve these issues - such as multi-factor authentication, blockchain for auditing, better encryption, and better machine-learning.
Imagine a data-sharing community for smart cities initiatives. Or a cybersecurity community with machine-learning from all threats to its members, not just threats against one member. Such a community has the potential to vastly improve its protection.
It's a matter of putting together all the right tools. We hope you'll start thinking about how you can help.
Victor Nilson - Senior Vice President - Big Data, AT&T
