NSF Grant Proposal Panel Summary

advertisement
Model-based Heavy Construction
Fischer & Peterson
NSF Grant Proposal Panel Summary
Proposal Number: 1000545
Performing Organization: Stanford University
NSF Program: Civil Infrastructure Systems
Principal Investigator: Fischer, Martin
Proposal Title: Model-based Integrated Heavy Civil Construction Project Monitoring
Objectives of the Proposal
The PI proposes a formal method for classifying project monitoring measurements. This is proposed as a
component of an integrated model-based project monitoring system for forecasting, monitoring, and
controlling project progress. The work builds on previous/ongoing research by the PI in the area of modelbased integrated construction, but focuses on the problem of measurement classification.
Intellectual Merit
The research addresses an important problem: automatic mapping of sensor-based measurements. However,
the scope of work and the research plan are lacking in clarity and purpose. The proposal lacks a scientific
approach/methodology - it suggests a four-step research work plan without proper explanation of the need,
expected outcomes, and connectivity between these steps. There is also no mention of other modeling and
mapping approaches (e.g. ontological approaches) that may offer a good solution to this problem of
mapping. The PI fails to discuss the work on construction ontologies and its relation to existing
classification approaches. In general, the proposal is poorly organized and is not well-written. Given, his
expertise in this area, the PI brings strength to the project. However, the chances of success are
questionable, since the scope and the proposed methodology are not clear.
Broader Impact
The PI is proposing the development of model-based project monitoring course material. There is no
planning for broadening the participation of underrepresented groups - there is no mention of any intent to
involve women or minorities. There is planning for the dissemination of research results through journal
publications and conference presentations. The proposer mentions that the categorization of measurements
would improve efficiency by 4% to 7% and that this benefits society directly as reduced waste of society's
resources. This is a big claim extremely difficult to validate. The broader impact on the society is unclear
and poorly justified.
Panel Summary Statement
The proposal addresses an important problem: automatic mapping of sensor-based measurements. However,
the scope of work is unclear, there is a lack of scientific approach/methodology, and the proposal is poorly
organized/written.
The summary was read by/to the panel and the panel concurred that the summary accurately reflects the
panel discussion.
Panel Recommendation: Do Not Recommend
2009 - 1000545
1
Rating
Intellectual merit
2009 - 1000545
Research Plan
Chance for
success
Review (summarized)
Broader impacts
The details of the classifying processes are This proposal is not well written, is poorly
missing. The proposal is well organized,
organized, is confusing and is lacking
presenting an excellent description of the
scientific rigor.
problem with a good bibliographical
research of relevant work but misses the
work on ontology and its relation to existing
approaches. Case and field validation is
proposed but does not test against existing
approaches, it’s not clear this approach is
best.
The proposal is full of war stories and
graduate student anecdotes, jargon, and is
very poorly organized and presented. Even
though the proposer argues the he will
publish his results, it is not all that clear what
the impact of these results will be beyond
their use in the CIFE classrooms at Stanford.
The objective of the proposed work is to
derive a method to classify the
measurements taken for model based project
monitoring. The proposal argues that such a
classification method would address the
problem of wasted resources and allow
better access of information to the workers
through a model based system.
While it is proposed to integrate the outcome
in course material, the impact on human
infrastructure and labs is not clear as stated.
However, the proposal describes efforts at
building partnerships with other agencies
and suggests dissemination through
publication. A generic classification for
project monitoring measurements would be
useful in improving productivity and
resource utilization.
The PI brings strength to the project based
on prior work in this area. However, the
chances of success are questionable since the
scope and the proposed methodology are not
clear.
On p12 the PI presents serious doubts about
the efficacy of the research.
The PI is proposing the development of
model-based project monitoring course
material, an improvement of efficiency by
4% to 7%, applied to infrastructure
benefiting society as reduced waste of
society's resources but this is a big claim
extremely difficult to validate. A plan for
publication is presented. No mention to
minorities.
The scope of work, the research plan and the While the proposal may enhance the
approach for analyzing the collected data
understanding of issues, it lacks scientific
lack clarity and purpose. The approach is
methodology
based on a four-step research work plan,
questionnaire surveys and case studies.
Proper explanation of the need, expected
outcomes, and connectivity between the
various steps.
The research plan is not innovative,
substantive, nor scientifically interesting.
Parts of the proposal are unreadable, for
example, the section "Plans not Used" p12,
the page numbers are missing and there are
typos.
The proposal is lacking in several aspects
related to the intellectual merit and broad
impact (as described above); and in general,
the proposal is not well-organized.
The PI intends to promote learning through
incorporation of results in coursework and
plans to disseminate results through
publication. Intent for broadening the
participation of underrepresented groups is
not discussed nor is justification provided for
the claims that auto-classification improves
efficiency from 4% to 7% or that reduced
waste benefits society.
There is no mention of other modeling and
mapping approaches (e.g. ontological
approaches) that may offer a good solution to
this problem.
The proposed formal method for classifying
project sensor-based measurements, as a
component of an integrated model-based
project monitoring system for forecasting,
monitoring, and controlling project progress,
is an important problem.
The novelty or intellectual merit of the
proposed project is questionable since
various formal and informal classification
systems exist (e.g., CSI, Revit, RS Means,
etc). However, the proposal argues that the
feedback of construction project progress is
unreliable and a generic classification of
project monitoring measurements would be
useful.
Through a model-based system for
classification of measurements, this proposal
seeks to improve project forecasting,
recording, and progressing, through reduced
errors and delays in communicating the
information. This is a worthy problem.
Good/Fair
4
The proposed work goal is to better
understand issues related to automatic
mapping of sensor-based measurements. The
PI is proposing to develop a formal process
for assigning classification to measurements,
a model-based classification filter, a modelbased automated classification, and a
method for mapping components within the
model-based system
Fair
3
Poor
2
Very Good/Good
Review 1
Model-based Heavy Construction
Fischer & Peterson
Summary
2
Model-based Heavy Construction
Fischer & Peterson
Review #1
Rating: Multiple Rating: (Very Good/Good)
REVIEW:
What is the intellectual merit of the proposed activity?
The proposed work goal is to better understand issues related to automatic mapping of sensor-based
measurements. The PI is proposing to develop a formal process for assigning classification to
measurements, a model-based classification filter, a model-based automated classification, and a method
for mapping components within the integrated model-based system
What are the broader impacts of the proposed activity?
The PI is proposing the development of model-based project monitoring course material, an improvement
of efficiency by 4% to 7%, applied to infrastructure benefiting society as reduced waste of society's
resources but this is a big claim extremely difficult to validate. A plan for publication is presented. No
mention to minorities.
Summary Statement
The PI claims that he has the intuition that the missing piece of the integrated control system, that this
research team is competent to address, is a method to classify the measurements but he doesn't clear
describes how he intends to do so. He is proposing a manual process, an effective method for automated
classification, and an internal mapping process but doesn't describes how he is planning to do this in details.
The PI describes the idea of applying Na´ve Bayes, k-nn, Rocchio, and SVM for classification but is not
clear that this is the best approach.
The proposal is well organized presenting an excellent description of the problem with a good
bibliographical research or relevant previous work but misses the work on construction ontology and its
relation to existing classification approaches.
The PI proposes validation trough case studies and field validation but there is no mention on
comparing/testing against other existing approaches.
2009 - 1000545
3
Model-based Heavy Construction
Fischer & Peterson
Review #2
Rating: Poor
REVIEW:
What is the intellectual merit of the proposed activity?
This proposal seeks to improve construction project progress monitoring in order to forecast, improve
record keeping, and control project progress so that reduced errors and delays in communicating the
information about the progress will results. This is a worthy problem. They seek to develop an integrated
model-based project monitoring system for improved classification of measurements.
The research plan is comprised of four phases: Phase I: Literature review; Phase II. Questionnaire
development; Phase III. Case Studies; Phase IV. Field trials. The research plan is not very innovative,
substantive, nor scientifically interesting.
Many parts of the proposal are unreadable. For example the following paragraph is presented under the
section "Plans not Used"(page 12): In the experimental aspect of the research a pure charrette was not used
in favor of including empirical 3rd party results. A quantitative analysis with-out case studies was not used
due to the typical small sample sets in construction research trials. Relying only on case studies was not
used since this would produce a toy research result."
There are not even page numbers in the proposal. There are also typos sprinkled throughout.
Finally, there are serious doubts written in the proposal about the efficacy of the research. On page 12,
under the heading Proposed activity consideration and organization: "The feedback from the heavy civil
industry is that this research is beyond anything they have on their table and is too uncertain to provide
significant resources beyond accommodating case studies and field experiments, contingent on nondisruption of business activities."
What are the broader impacts of the proposed activity?
The proposal is full of war stories and graduate student anecdotes, jargon, and is very poorly organized and
presented. Even though the proposer argues the he will publish his results, it is not all that clear what the
impact of these results will be beyond their use in the CIFE classrooms at Stanford.
Summary Statement
This proposal is not well written, is poorly organized, is confusing and is lacking scientific rigor.
2009 - 1000545
4
Model-based Heavy Construction
Fischer & Peterson
Review #3
Rating: Fair
REVIEW:
What is the intellectual merit of the proposed activity? Is the
proposed science Novel?
The novelty or intellectual merit of the proposed project is questionable since various formal and informal
classification systems exist (e.g., CSI, Revit, RS Means, etc). However, the proposal argues that the
feedback of construction project progress is unreliable and a generic classification of project monitoring
measurements would be useful.
Is there a coherent research plan?
The scope of work and the research plan are lacking in clarity and purpose. The proposal suggests four step
research work plan without proper explanation of the need, expected outcomes, and connectivity between
the various steps.
What are the technical strengths and weaknesses of the approach?
The approach is based on questionnaire surveys and case studies. The approach for analyzing the collected
data is not clear as stated.
What are the chances the research will succeed? Are the PIs qualified
to do the research?
The PI brings strength to the project based on his prior work in this area. However, the chances of success
are questionable since the scope and the proposed methodology are not clear as stated.
What are the broader impacts of the proposed activity?
Will the work be reflected in education programs?
It is proposed to integrate the outcome in CIFE course material and lab assignments.
Will it help build human infrastructure, labs and partnerships for
research enterprise?
The impact on human infrastructure and labs is not clear as stated. However, the proposal describes efforts
at building partnerships with other agencies and NIST.
Will it be disseminated effectively?
The proposal suggests dissemination through standard publication outlets such as journals and conferences.
How will the research impact the economy, the environment, social
well being and the general evolution if it succeeds?
A generic classification for project monitoring measurements would be useful in improving productivity
and resource utilization.
Summary Statement
The objective of the proposed work is to derive a method to classify the measurements taken for model
based project monitoring. The proposal argues that such a classification method would address the problem
of wasted resources and allow better access of information to the workers through a model based system.
2009 - 1000545
5
Model-based Heavy Construction
Fischer & Peterson
Review #4
Rating: Multiple Rating: (Good/Fair)
REVIEW:
What is the intellectual merit of the proposed activity?
The PI proposes a formal method for classifying project monitoring measurements. This is proposed as a
component of an integrated model-based project monitoring system for forecasting, monitoring, and
controlling project progress. The work builds on previous/ongoing research by the PI in the area of modelbased integrated construction, but focuses on the problem of measurement classification. The research
addresses an important problem: mapping/classifying sensor-based measurements. The proposal may
enhance the understanding of issues related to automatic mapping of sensor-based measurements. However,
the proposal lacks scientific methodology. There is also no mention of other modeling and mapping
approaches (e.g. ontological approaches) that may offer a good solution to this problem.
What are the broader impacts of the proposed activity?
The proposer intends to promote teaching and learning through incorporation of research results in
university coursework. There is no planning for broadening the participation of underrepresented groups û
there is no mention of any intent at all. There is planning for the dissemination of research results through
journal publications and conference presentations. The proposer mentions that the categorization of
measurements would improve efficiency by 4% to 7% and that this benefits society directly as reduced
waste of society's resources. However, there is no justification for these claims.
Summary Statement
The proposal is lacking in several aspects related to the intellectual merit and broad impact (as described
above); and in general, the proposal is not well-organized.
2009 - 1000545
6
Model-based Heavy Construction
Fischer & Peterson
Context Statement
This year the Division of Civil, Mechanical and Manufacturing Innovation expects to review about 3900
competitive research proposals, and expects to make awards to between 20% and 25% of them.
The Division's practice is for programs to seek the advice of several independent reviewers for each
proposal, and these reviewers and reviewers for other proposals submitted to the program, comprise a panel
to compare and assess the merit of related proposals. The panel prepares a summary of its discussion on
each proposal. Your proposal was recently considered by the Civil Infrastructure Systems Unsolicited
The Panel Summary and verbatim copies of all completed reviews are available via FastLane. In reading
them, please keep in mind that reviewers are addressing their comments primarily to the NSF, not
necessarily to you. They sometimes make remarks without giving detailed references or providing specific
suggestions for improvement, although many reviewers do provide such helpful information. Some reviews
may contain non-substantive, irrelevant or erroneous statements that the Program Director did not use in
making a recommendation.
Decision about a particular proposal is often very difficult, and factors other than reviewer comments and
ratings enter into the decision. Comments by a reviewer must sometimes be considered in the context of
other reviews by the same person. A Program Director often has additional information not available to
reviewers (such as progress reports on recent projects). Maintaining appropriate balance among subfields,
the availability of other funding, the total amount of funds available to the program, and general
Foundation policies are also important decision factors.
Additional information about the decision on your proposal is provided separately. If more information is
needed, please contact the cognizant Program Director.
Information about reconsideration of declined proposals is found in NSF's Grant Policy Manual
http://www.nsf.gov/pubs/policydocs/pappguide/nsf10_1/gpg_4.jsp#IVD
2009 - 1000545
7
Download