THE INTERNET OF THINGS, PRODUCTIVITY AND EMPLOYMENT Boston 0915
1. THE INTERNET OF
THINGS, PRODUCTIVITY
AND EMPLOYMENT
Robert B. Cohen, PhD,
Director, Project on the Economics and Business Impacts of the New IP
and Internet of Things
Economic Strategy Institute www.econstrat.org
Presentation at Internet of Things Summit, Boston, Sept. 8-9, 2015
2. Major Impacts of IoT
1. Implied Efficiencies in Operations and
Possible Productivity Gains
2. Significant Buildout of Infrastructure
3. Important Change in Employment
-- New jobs merge skills drawn from earlier “siloed” occupations: DevOps; “Cross-skilling”
-- Machine learning and new tools to handle additional IoT protocols automate massive
sensor networks.
3. Milestones in the Growth of IoT
Stan Schneider, “Understanding The Protocols Behind The Internet Of Things,”
Electronic Design, October 9, 2013. http://electronicdesign.com/iot/understanding-protocols-behind-internet-things
4. What does DevOps Contribute
1. Ops groups let Developers manage
the “operational characteristics” of
apps they are building
2. Developers must be mindful of
Operational considerations when they
build apps. In the DevOps world this is
a ‘Shift Left’.
3. “As we move towards Software
Defined Environments, we have the
ability to build, version and manage
complex environments, all as code.”*
*Sanjeev Sharma, “Adopting DevOps – Part III: Aligning the Dev and Ops Teams,” May 9, 2013,
https://sdarchitect.wordpress.com/2013/04/12/adopting-devops-part-iii-aligning-the-dev-and-ops-teams/
**Steve Hall, “DevOps Requires New Job Skills & Roles,” https://www.scriptrock.com/blog/devops-new-job-skills-roles-titles
DevOps Automation Engineer is to analyze, design, implement and
validate strategies for continuous deployment
Release managers manage software released from development
stage to software release
Cloud architect constructs meta-architecture - fields of “resources
upon which applications are dynamically developed and deployed”
Integration specialist / test automation specialist enables the
“DevOps team to integrate and build code regularly and
efficiently”**
5. The Engineers’ Dilemma and How It
May be Solved
1. “Enterprise networks require an engineer to support some 150 network devices” –Nick
Lippis, Open Networking Users’ Group (ONUG)
2. With 50 Billion connected devices feeding into enterprise networks, that would require
about 330 million engineers. You can see where this is going. That is clearly not going to be
possible providing support as we do in today’s world!
3. So what’s going to happen?
….. Software Defined Networking becomes a way “to achieve much greater automation of
network management”
….. Programming Protocol-Independent Packet Processors (P4) lets “the controller … tell the
switches how the controller wants them to act,” as “IoT will lead to new network protocols,
the protocol-independence of P4 will be …hugely valuable.”*
* Personal communication with Prof. Jennifer Rexford of P4.org. Also see Jennifer Rexford with others,
“Programming Protocol-Independent Packet Processors,” http://arxiv.org/abs/1312.1719
6. Economic Efficiencies
1. Sensor Networks save money in maintenance costs: GE and aircraft engines
2. eHealth Networks reduce the need for care providers and hospitals: shift focus to data
This may depend on whether there is an offsetting demand effect, if people seek more care and better
levels of care, resulting in greater demand from the health care system.
3. Do the economic efficiencies result in productivity gains?
If health care networks operate more efficiently and more treatment is provided per hour or per worker,
productivity gains may be important.
Finland has estimated that a focus on IoT may result in raising national productivity from 2.5% to 3.9%
and keeping economic growth steady in spite of a population decline.
7. IoT Provides a Boost to Productivity and GDP
VTT Technical Research Centre of Finland, “Productivity Leap with IoT: Visions of the Internet of Things with
a special focus on Global Asset Management and Smart Lighting,” http://www.vtt.fi/publications/index.jsp
8. Phases for Impacts of IoT
1. Big Data and Analytics Emphasized
2. Massive Sensor Networks for Driverless Cars, Monitored Patients
Greater use of Containers to manage widely dispersed networks
with greater processing at the periphery of new networks – link to
“Fog Computing.”
Increased automation of lower-value processes in computing,
storage and networking.
3. Interactive Sensor Networks that control the action of connected
devices, whether Smart Grids or Wearable Health Devices
9. Orchestrating Containers
Greg DeMichillie, “In case you missed it in January: Cloud pricing, complex business challenges in the cloud and our new
series unpacks containers,” Google Cloud Platform Blog, Jan. 30, 2015.
http://googlecloudplatform.blogspot.com/2015_01_01_archive.html
10. Employment Impacts in Different Phases
of IoT
Phase 1. Increased Demand for data analysts
↑ Jobs evaluating data from sensor networks
Phase 2. Increased Infrastructure Build-out and Increased Demand for Managing
Infrastructure, Storage and Processing at the Edge of Networks. More analysis at the
Edge. See Cisco’s Fog Computing papers -
http://www.cisco.com/web/solutions/trends/iot/fog-computing.html
Phase 3. Upgrading Infrastructure to be more “two-way” or with complex
connectivity – Management of Intelligence; more employees to manage and evaluate
data
11. New “Cross-Skilled Jobs”
There is a general trend to “re-skilling” jobs, adding additional requirements to traditional occupations and
requiring a broader range of skills in DevOps and Continuous Service Delivery and Microservices teams.
1. To run early networks, data analysts need to also know something about networks, security, sensors.
Data analyst jobs are strong in analytic skills but also require software skills working with various Apache tools, Mesos and
MapReduce.
2. DevOps groups building new services and apps for analysis and for customers include software developers, software testers
and software deployment experts.
3. New jobs such as “Platform Engineers” working in Continuous Service Delivery and Microservices teams combine a diverse
range of skills for software development, such as software creation, network expertise, and data storage expertise.
12. Continuous Service Delivery: Product Teams
Design teams replace isolated
“skill areas,” like quality
assurance; drastically reduce the
number of steps to create
software
Impact: Better knowledge
of customers, new sales
opportunities.
Adrian Cockcroft, “Creating Business Value with Cloud Infrastructure,” Open Networking User Group, May 13-14, 2015.
13. Software and Computer Applications Career Ladder
Help Desk / Entry
Level Support
Experience
Jr / Mid Level Software
Developer
Senior Software
Developer
Mobile
Applications
Developer
Advanced Computer
Support
Systems Analyst
BI Architect / Engineer
Jr / Mid Level QA
Tester / Engineer
Senior QA Tester /
Engineer
Computer Programming Internship
BI Analyst
Burning Glass Technologies, “Burning Glass Technologies Opportunities in Chicago’s IT Landscape.”
14. Job Gains and Losses in IoT: Data
Analytics and Driverless Cars
Healthcare: Joining Genomic Data and Patient Care Data
↑ Data Analysis and Data Management Jobs
↓ MD, Nursing and health support staff jobs
↓ Health Care costs, so
↑ Consumer Surplus
Massive Sensor Nets for Driverless Cars
↑ ↑ Infrastructure construction jobs (First of several likely waves of building)
↑ Sensor network management jobs
↓ Emergency response jobs
↓ ↓ Spending on cars due to shift to “Cars as a Service” with fewer jobs in auto firms and their
suppliers but
↑ Jobs in transport sensor systems management, installation and sales
↑ Consumer Surplus (overall cost of using transportation is lower)
,
15. Automation, Machine Learning and IoT
1. CEO of BNY Mellon links the use of advanced IT systems to machine learning, saying
machine learning can be used to “eliminate a lot of repetitive work that isn’t exciting for
people to do and doesn’t add a lot of value to the business.”
2. Software Defined Networking becomes a way “to achieve much greater automation of
network management.” One route is via Programming Protocol-Independent Packet
Processors (P4) that let “the controller to tell the switches how the controller wants them to
act,” as “IoT will lead to new network protocols, the protocol-independence of P4 will be
…hugely valuable.”*
3. Jobs may eliminate much repetitive and low value work, become more creative and provide
more direct value. This outcome from automation and machine learning differs sharply from
what is proposed in Frey and Osborne* and in The Second Machine Age.
Carl Benedikt Frey and Michael A. Osborne, “The Future Of Employment: How Susceptible Are Jobs To Computerisation?”
16. Takeaways
1. Process and architectural changes related to IoT will bring about new efficiencies
and raise productivity and output. Without this type of progress, it will be difficult to
operate massive networks with about 50 billion connected devices.
2. Software Defined Environments will provide the ability to build, version and
manage complex environments, all as code
3. New Jobs will be “cross-skilling” jobs that will merge functions that were previously
more “siloed.” New jobs will also be created in teams that deliver software or manage
operations more efficiently. Both types of positions will benefit from machine learning
and from the ability to work with new protocols by using approaches like Programming
Protocol-Independent Packet Processors (P4).