10. Risks and Risk Mitigation
10.1 Politics
Risk: GS-21 must be a technical effort guided by an
impartial government entity (OAI?). If GS-21 takes on the appearance of a
partisan political effort, it will die from distraction. In that regard, any
appearance of crony capitalism or congressional or activist interference could
kill the effort.
Mitigation: GS-21 will require high cover from the White
House to make clear, at every challenge, that this effort is meant to 1)
improve working conditions for government workers, 2) improve service to the
American public, and 3) save taxpayer money and therefore needs to be allowed
to succeed. GS-21 decisions must be based on technical and cross-agency programmatic
criteria. Political decisions will be very difficult to defend.
10.2 Internal resistance
Risk: Workers, agencies, and users all dislike change.
Resistance to GS-21 can be expected from all directions.
Mitigation: OAI must make users an integral part of the
development process.
10.3 Skepticism
Risk: On one hand, there have been failures. On the other
hand, some are terrified that, even if OAI succeeds, someone may benefit
unfairly.
Mitigation: Impartiality, transparency, and a careful
incremental approach are necessary.
10.4 Impatience
Risk: Big projects tend to allocate too little time to Team
building, Planning, Socialization, Requirements definition, Architecture definition,
and Standards specification. During design
·
too little attention is given to Community
feedback, Comment review, and Response,
·
too little time is left for Alpha and Beta test
and refinement, and
·
there is a temptation to take on too much at
once – too many layers, agencies, or activity domains.
Mitigation: Celebrate successes and plan following steps
deliberately.
10.5 Leadership
Risk: GS-21 is an enormous undertaking. Project leaders may
be
·
Slow to make vital decisions,
·
Unable to resolve team disagreements,
·
Unable to motivate the team on behalf of program
sponsors,
·
Unable to defend the team and the program to program
sponsors.
Mitigation: GS-21 will need clear goals and a team-oriented,
apolitical, disinterested leader with the courage to take risks, admit
mistakes, and credit successes to the team.
10.6 Funding
Risk: Money will be closely watched. Underfunding could
choke off vital operations; and overfunding could breed waste and make the
effort a target for cancellation.
Mitigation: GS-21 should start with a “bare bones” budget,
build the team, and request more funding as needed. Since agencies have a stake
in GS-21 success, cost sharing has the potential to significantly reduce
funding risk.
10.7 The Roadmap
Risk: The schedule may not be quite right. If it is too
ambitious, the team may miss deadlines and look bad. If it is too cautious, the
work may slow to meet expectations.
Mitigation: The Roadmap must be a living document, kept
current as the effort proceeds.
10.8 Technical setbacks
Risk: Not everything will go as planned. GS-21 represents
innovation and new development.
Mitigation: make the ITE Facility a Center of Excellence,
open for critical experiments (needed to mitigate technical risk) and for demonstrations
on short notice – keep it Great.
10.9 Innovation never works in government
Risk: Even when they spend millions of dollars, DARPA and
ONR rarely transition anything. Although this is not entirely fair, those
agencies generally rely on others to make transition happen.
Mitigation: Invest in and insist on high-quality system
engineering. The innovation must be proscribed “within the layers,” observant
of standardized interfaces
10.10 Security, security, security
Risk: the external threat ranges from mischief to crime and
espionage; the internal threat represents a violation of trust and presents a
related, but separate set of challenges.
Mitigation of external threat: Inconvenience – maximum
convenience results in maximum risk – there must be tight, well-protected
pathways to data; data must be accessible, but access must be tightly
controlled. Defense in depth must include physical protection, partial
isolation, and trusted processes and personnel. The ITE Facility must be
accessible to critical security experiments, where live data are not exposed.
Security measures must themselves be protected (classified). Inconvenience
raises costs for all: system developers; authorized users; and unauthorized
users.
Mitigation of internal threat: verification of insider
trustworthiness can help, but as we have learned from double agents, is not
foolproof. Access logs, redundant cross checks, data and access partitions,
rotation of admin privileges, and (admin) term limits can help, but the insider
threat will continue to be a difficult challenge.
Mitigation of data loss/destruction threat: continual data
backup; geographic distribution of data; a transaction paradigm; and
elimination of single points of vulnerability.
R&D: investment in detection. ID, pursuit, and prosecution
to raise the cost to hackers.
