The Apache Software Foundation (ASF) recently announced that SINGA, a framework for distributed deep-learning, has graduated to top-level project (TLP) status, signifying the project’s maturity and stability. SINGA has already been adopted by companies in several sectors, including banking and healthcare.

Originally developed at the National University of Singapore, SINGA joined ASF’s incubator in March 2015. SINGA provides a framework for distributing the work of training deep-learning models across a cluster of machines, in order to reduce the time needed to train the model. In addition to its use as a platform for academic research, SINGA has been used in commercial applications by Citigroup and CBRE, as well as in several health-care applications, including an app to aid patients with pre-diabetes.

The success of deep-learning models has been driven by the use of very large datasets, such as ImageNet with hundreds of thousands of images, and complex models with millions of parameters. Google’s BERT natural-language model contains 300 million parameters and is trained on nearly 3 billion words. However, this training often requires hours, if not days, to complete. To speed up this process, researchers have turned to parallel computing, which distributes the work across a cluster of machines. According to Professor Beng Chin Ooi, leader of the research group that developed SINGA:

It is essential to scale deep learning via distributed computing as…deep learning models are typically large and trained over big datasets, which may take hundreds of days using a single GPU.

There are two broad parallelism strategies for distributed deep-learning: data parallelism, where multiple machines work on different subsets of the input data, and model parallelism, where multiple machines train different sections of the neural-network model. SINGA supports both of these strategies, as well as a combination of the two. These strategies do introduce some communication and synchronization overhead, required to coordinate the work among the machines in the cluster. SINGA implements several optimizations to minimize this overhead.

Acceptance as a top-level project means that SINGA has passed several milestones related to software quality and community, which in theory makes the software more attractive as a solution. However, one possible barrier to adoption is that instead of building upon an existing API for modeling neural networks, such as Keras, SINGA’s designers chose to implement their own. By contrast, the Horovod framework open-sourced by Uber allows developers to port existing models written for the two most popular deep-learning frameworks, TensorFlow and PyTorch. PyTorch in particular is the framework used in a majority of recent research papers.

ASF has several other top-level distributed-data processing projects that support machine-learning, including Spark and Ignite. Unlike these, SINGA is designed specifically for deep-learning’s large models. ASF is also home to MXNet, a deep-learning framework similar to TensorFlow and PyTorch, which is still in incubator status. AWS touted MXNet as its framework of choice in late 2016, but MXNet still hasn’t achieved widespread popularity, hovering at just under 2% in KDNugget’s polls.

Apache SINGA version 2.0 was released in April, 2019. The source code is available on GitHub, and a list of open issues can be tracked in SINGA’s Jira project. According to ASF, upcoming features include “SINGA-lite for deep learning on edge devices with 5G, and SINGA-easy for making AI usable by domain experts (without deep AI background).

The post Deep-Learning Framework SINGA Graduates to Top-Level Apache Project appeared first on Artificial Intelligence.

Artificial intelligence (AI) everywhere, transparently immersive experiences and digital platforms are the trends that will provide unrivaled intelligence, create profoundly new experiences and offer platforms that allow organizations to connect with new business ecosystems, according Gartner’s Hype Cycle for Emerging Technologies 2017.

“In addition to the potential impact on businesses, these trends provide a significant opportunity for enterprise architecture leaders to help senior business and IT leaders respond to digital business opportunities and threats by creating signature-ready actionable and diagnostic deliverables that guide investment decisions,” said Mike J. Walker, research director at Gartner.

Artificial intelligence technologies will be the most disruptive class of technologies over the next 10 years due to radical computational power, near-endless amounts of data, and unprecedented advances in deep neural networks; these will enable organizations with AI technologies to harness data in order to adapt to new situations and solve problems that no one has ever encountered previously.

Enterprises that are seeking leverage in this theme should consider the following technologies Deep Learning, Deep Reinforcement Learning, Artificial General Intelligence, Autonomous Vehicles, Cognitive Computing, Commercial UAVs (Drones), Conversational User Interfaces, Enterprise Taxonomy and Ontology Management, Machine Learning, Smart Dust, Smart Robots and Smart Workspace.

Also, technology will continue to become more human-centric to the point where it will introduce transparency between people, businesses and things. This relationship will become much more entwined as the evolution of technology becomes more adaptive, contextual and fluid within the workplace, at home, and in interacting with businesses and other people.

Critical technologies to be considered include 4D Printing, Augmented Reality (AR), Computer-Brain Interface, Connected Home, Human Augmentation, Nanotube Electronics, Virtual Reality (VR) and Volumetric Displays.

Further, emerging technologies require revolutionizing the enabling foundations that provide the volume of data needed, advanced compute power, and ubiquity-enabling ecosystems. The shift from compartmentalized technical infrastructure to ecosystem-enabling platforms is laying the foundations for entirely new business models that are forming the bridge between humans and technology.

Key platform-enabling technologies to track include 5G, Digital Twin, Edge Computing, Blockchain, IoT Platform, Neuromorphic Hardware, Quantum Computing, Serverless PaaS and Software-Defined Security.

The post 3 megatrends for digital business into the next decade appeared first on Artificial Intelligence.

Monsanto CIO Jim Swanson has outlined five key elements of Monsanto’s digital transformation: customer centricity, internal business process disruption, technology and automation, data and decision science, and leadership and change management.

He views all of those facets as integral to breaking through the “clay layer” of mid-level employees, present in every organization, who may resist short-term changes in spite of long-term gain — the rank-and-file who can either inhibit digital transformation or become its greatest champions. In so doing, he asserts, IT is poised to take on its most rigorous and most critical role of “transformation agent,” providing tangible and sustainable value while moving the business forward.

4 pillars of Monsanto’s IT operating model

Swanson’s IT operating model is grounded in four key pillars.

The first is modernization to ensure business continuity and stability, which at Monsanto includes such initiatives as the adoption of a hybrid cloud environment and network improvements to help ensure high-volume data transfer on a global scale.

The second pillar is data, the raw material underpinning Monsanto’s data models: The IT department has created a data abstraction layer supported by application programming interfaces (APIs) to help employees consume and synthesize internal data quickly.

A third strategic pillar hinges on the six major platforms — from field operations to finance — whose attendant products and capabilities are expanded upon in an iterative fashion.

The fourth pillar of Monsanto’s IT operating model is data science.

“We embed data science in all that we do — in our products, in our platforms, in the data — so every key decision that we make . . . [has] a data science model,” Swanson explains.

Data science at Monsanto is embedded in sales account plans so that account representatives are better informed about customer priorities and information needs. And, critically, Swanson explains that data science is a core component of the supply chain and logistical processes that drive Monsanto’s business.

“We are digitizing our whole supply chain,” he says. “And we call it a connected supply chain, from the farmer growing [the crop] to our manufacturing plants that are packaging and drying it, to our distribution to the farm gate.”

Using data science to reinvent the supply chain

Monsanto’s research and development (R&D) department was very forward-thinking when it came to data science, Swanson explains, “but that same energy wasn’t appearing in supply chain and commercial.” Swanson had a strong partner in supply chain, and he worked with his colleague to talk through how Monsanto’s supply chain could transform.

One of the significant issues to overcome was seed loss. A single Monsanto seed truck contains tens of thousands of dollars’ worth of product. During the journey from the farm to the manufacturing plant, these trucks can queue for miles, becoming extremely hot in the process. Extreme heat can negatively impact germination potential, and Swanson sought to compensate by implementing sensors on the back of the trucks.

+ Related: Turning data into food +

“We put sensors in back of these trucks that [measured] temperature, pressure, geolocation, and we streamed that data through IoT to the plant managers through a dashboard,” he explains. “And they got to see every truck, where it was in their geography, where it was on track to be able to be offloaded, and if it was in the queue three miles back before it’s offloaded — what’s the temperature of that truck?”

The sensors allowed Monsanto to prioritize which trucks made it into the manufacturing plant first, providing a streamlined and data-driven set of efficiencies to the supply chain.

The connected supply chain continues within the manufacturing plant, Swanson says.

“The plant workers work on iPads now, looking at sensor data around their equipment, looking at plant logistics, looking at how to optimize the flow of product in the plant via tools and data they’ve never had before,” he says. “And they can optimize that flow, reduce cost of goods, and get a higher [quality] product out to the marketplace to our distributors and dealers.”

Bridging the data divide with farm distributors and customers

Data science at Monsanto extends to the most crucial interactions that Monsanto has with its customers. An issue that continues to impact Monsanto’s profit and loss (P&L) statement is that of product returns, specifically seed returns. Growers, seeking to optimize their yield, would select different seed varieties at the beginning of the growing season, such as corn and soybeans, and ultimately return seeds they did not plant to Monsanto.

To make an estimate of projected losses due to returns, Monsanto account representatives would call growers weekly.

“There are two problems with that,” Swanson says. “One, it’s not very value-add for our customer — a rep calling just asking, ‘Are you returning seeds? We’re trying to calculate at the end of the year how it is going to impact us.’ Secondly, time with the grower is precious for a sales rep. They’d much rather be detailing a product or an opportunity than call them about how many seeds they’re going to return.”

Monsanto IT applied a data model to the issue. Swanson reports that the resulting model provided a high degree of efficacy in predicting returns, providing months of projections and, by extension, lead time. His strategy has been to reinforce such modeling in arenas as diverse as the effectiveness of marketing campaigns and customer segmentations, while following a consistent procedure: Optimize data inputs, test, iterate, scale and then educate colleagues and customers.

Data science, Swanson explains, has been employed to help maximize agricultural outputs while minimizing inputs (water, fertilizer, seed). Arable land is an inherently finite resource, and Monsanto’s data science models have helped growers and distributors alike plot efficiencies.

Monsanto has identified approximately 40 major decisions a farmer makes every year, from how much water to plot to which agronomic practices produce the most resilient plant phenotypes. Monsanto harvest advisors employ the company’s Climate platform, Swanson continues, and use data to provide consultative guidance to growers. They analyze, for example, nitrogen leakage in the soil from fertilizer, as well as how to minimize the input (fertilizer) while maximizing crop output. The company has advised European farmers on efficiencies in water irrigation practices and farmers in India using its Farm Rise platform — allowing growers on the subcontinent to discern when to optimally sell such commodities as wheat.

The rise of the Digital Outreach Group

Swanson uses the term “digital radar” to describe the evolving set of capabilities that Monsanto rigorously tests in the pursuit of its ongoing transformation. Because IT helps support an integrated view of customer value and internal processes based upon a variety of data points, it is imperative that Monsanto’s IT division help propel conversations on the organization’s “digital radar,” he says.

To this end, Swanson helped organize the Digital Outreach Group, a coterie of MBAs and scientists who discuss the organization’s digital footprint and relevant potential investment targets. The Digital Outreach Group meets with more than 250 startups every year and approves approximately 30 proof of concepts (POCs). Among those POCs, Swanson says, perhaps five new technologies will be adopted within the company — from IoT to cybersecurity. The Digital Outreach Group collaborates closely with Monsanto Growth Ventures, the venture capital arm of the company that specializes in equity investments in agriculturally focused organizations.

IT leaders must ask how technology and data can help transform processes

Swanson strongly emphasizes that IT leaders must take an assertive approach to solving business problems and integrating customer solutions. At Monsanto, this is partly a function of governance. The CIO shares that he pivoted the organization to allow for a “thin layer” of IT leaders who sit on leadership teams, interfacing directly with cross-departmental colleagues. This structure allows teams, Swanson asserts, to ask how technology and data can transform what employees do, not just incrementally improve it.

Success, he continues, is also partly a function of process. He says organically embedding relevant data insights into Monsanto’s day-to-day employee functions helps employees adopt new processes easily because they see the value; it is a matter of breaking through the “clay layer” of potential distraction and ambivalence by making employees’ lives easier.

Swanson explains:

“If I ask [salespeople] to go run a model, to go figure out who they should be spending more time with among their customers, they are never going to run the model. If I embed it in their CRM, and they do their account plan, and behind the scenes I’m running the model . . . and their account plan just [states], ‘You should be spending this amount of time with Farmer Joe, and this is what they need,’ it’s part of their workflow. And when they see [data as] part of their workflow, they can actually iterate on it because it’s part of what they do every day.

“That’s really the secret sauce that we’ve been able to implement, to create this iteration that we need around the data constantly, and expose why [employees] need to care about data because it directly impacts their ability to do their job.”

Finally, Swanson says, the journey is partly a function of culture. He says a diverse array of viewpoints and skill sets are necessary, as they “bring together the breeder, the commercial person, the marketer, the supply chain lead, the IT person, [and] the finance lead to solve a problem or create an opportunity.”

Swanson speaks regularly with human resources about how to better embed technology and data skill sets to every single role in the company, as well as with his colleagues on how to inculcate a “test and learn” culture where iteration is the norm.

“People shouldn’t be thinking of IT as some back-office function that you call when there is just a help desk issue,” he says. “They really should be looking at it as part and parcel of driving substantial value for companies that are really willing to use IT as a growth driver versus just an enabler or an operationally efficient function.”

The post Data science key to Monsanto improving its supply chain appeared first on Artificial Intelligence.