Assignment Prompt & Instructions
– 2 years ago
**files (1)**
– ## Case.pdf
**## Case.pdf**
2
Please Note: The fictional case provided herein was
developed over the past seven months, before the current
California fires. Lockheed Martin is deeply committed to
supporting the relief and recovery efforts of those affected by
the fires. We have donated to support immediate relief and
long-term recovery of communities impacted by the Los
Angeles wildfires. You can learn more about our efforts here.
We understand that the theme of firefighting intelligence is
particularly relevant but also sensitive given the current
situation. We want to acknowledge that Lockheed Martin
products are being used to assist in the firefighting efforts in
Los Angeles. While our technology is designed to support
these efforts, we do not want to draw attention away from
the very real challenges and tragedies that are unfolding in
California. Our intention is to provide a thought-provoking
and educational experience for all participants, and we hope
that this case will serve as a valuable learning tool.
[https://www.lockheedmartin.com/en-us/news/statements-speeches/2025/los-angeles-wildfires-relief-and-recovery.html](https://www.lockheedmartin.com/en-us/news/statements-speeches/2025/los-angeles-wildfires-relief-and-recovery.html)
3
2025 Ethics in Engineering Case
Background
In 2023, there were over 56,000 wildfires across the US burning 2.6 million acres of land
and destroying over 4,300 structures including around 3,000 homes. The total cost of
these wildfires is estimated to be between $394 to $893 billion each year.
Analyzing wildfire intelligence information and deploying firefighting resources is a time-
consuming process that requires multiple manual decisions. This is a complex process
and due to time and resource constraints, involving humans in every decision is inefficient.
Because of the changing nature of wildfires, the current process is expensive and prone to
human error. The use of advanced technologies, including artificial intelligence (AI),
machine learning (ML), and other automated methods, could significantly reduce data
processing times and improve data accuracy. This would enable incident commanders,
division supervisors, and hand crew bosses to make critical decisions more rapidly,
thereby reducing risk to life, property, structures, and land.
To modernize the approach, reduce cost, and better serve those in high fire risk areas, the
National Interagency Fire Center (NIFC) is assessing new technologies to better manage
resources during wildfires. They are developing a firefighting intelligence prototype
solution called FIREINT (Firefighting Intelligence), which will be used to assess fire
conditions, analyze the environment and location of the fire, and determine the most
efficient use of resources, including human intervention in verification and decision
making, to combat the fire to best protect the surrounding area. The development of
FIREINT is expected to be completed by May 2025, with plans for the operational system
to be deployed prior to the peak of the 2026 wildfire season.
The objectives of FIREINT are to:
• maximize the use of technology while ensuring human involvement in critical decisions.
• assist incident commanders in prioritizing the deployment of scarce resources while
providing maximum protection to emergency responders.
• enable developers to easily integrate and swap out various AI algorithms, data sources,
and interfaces, allowing for rapid customization and adaptation to changing wildfire
response scenarios.
The Capture
The NIFC has contracted three companies to develop a prototype FIREINT solution to be
demonstrated to them in May 2025.
The NIFC has specified the requirements below within which the FIREINT demo shall
operate. A representative map highlighting key features associated with the demo is
included in Figure 1.
4
FIREINT shall:
1. be capable of assessing wildfires in an area of up to 200,000 acres
(approximately 300 square miles) with various terrains and population zones
found in the western and southwestern US to inform incident commanders
where to deploy resources.
2. notify when human decision making is required to minimize risk to human life
and resources.
3. base fire suppression solution on deployment of available ground based and
aerial resources. This consists of 23 air tankers, 11 transport helicopters, and 14
tactical coordination aircraft for fire reconnaissance. Other types of resources
and technology to support traditional firefighting resources can be part of the
solution.
4. provide near real-time information to incident commanders and emergency
responders.
5. explain the decision-making processes for when to use different levels of
technology and human interaction while taking into account the following
considerations.
a. Protection and safety of firefighters;
b. Prioritization of human lives versus property;
c. Protection of critical infrastructure needed to support homes and life;
d. Evaluation of environmental impact from both the fire and suppressant;
e. Optimization/preservation of fire suppression asset utilization; and
f. Prioritization of monetary and historic value impact.
Companies will be contracted up to a value of $1M for the prototype solution which
includes the demonstration. Following the demonstration of each company’s
solution, the NIFC will select the winning solution and award a follow-on contract
valued at $10M.
HotShot Response Inc. (HotShot)
HotShot is one of the three companies selected by the NIFC to develop the
prototype FIREINT solution by May 2025. HotShot has a long history of providing
exceptional support to the NIFC. They have a record of consistently achieving the
mission to significantly reduce casualties during high risk and constantly changing
fire and weather situations via their logistics and firefighting asset deployment
systems. This mission is deeply embedded into the culture of the company. The
existing architecture's limited openness and finite scalability hinder timely upgrades,
underscoring the need for a modular, open-source system that can seamlessly
integrate with diverse components and adapt to evolving wildfire response
requirements.
5
HotShot’s president, Kevin O’Sullivan, realized months before this opportunity came up,
they needed to modernize their current product line and develop discriminating capabilities
to remain a leading partner for the NIFC. He felt they had internal expertise to advance
automation, but they needed external resources to incorporate AI capabilities.
HotShot’s chief engineer, Kimberly Miller, is not in total agreement with needing AI to
advance their solution and continues to fight against the need for using it. They have used
their current strategy for the last 25 years with great success without needing outside help.
O’Sullivan respects Miller’s viewpoint but sees the future technology landscape changing
and recognizes the need to integrate automation and AI into their current solutions. As a
result, O’Sullivan entered negotiations to acquire DeciderAI.
DeciderAI
DeciderAI is an innovative company with a proven and fielded AI resource deployment and
logistics solution that addresses hurricane emergency management, but they lack the
resources to expand into adjacent markets. DeciderAI leverages a suite of key enabling
technologies to optimize the deployment and allocation of resources in complex, dynamic
environments. At the core of their platform lies a robust combination of AI algorithms,
which enable real-time predictive analytics, pattern recognition, and decision-making. Their
proprietary graph database management system provides a scalable and flexible data
structure for modeling and analyzing complex networks, allowing them to identify resource
limitations and predictive schedule conflicts.
DeciderAI’s Chief Engineer, Hunter Garcia, has been a pioneer in the field of AI for almost
20 years and, based on what he knows, he can develop a system with little to no human
intervention. He feels AI is smarter than any human and can take all bias out of the
decision-making process. Garcia doesn’t have firefighting intelligence experience but can
easily see how existing AI processes could be applied. In contrast to HotShot’s chief
engineer, Garcia feels the strategy and solutions should always be evolving to incorporate
the latest technologies. Those who stagnate get left behind.
HotShotAI – Acquisition of DeciderAI by HotShot
On December 31, 2024, HotShot successfully completed its acquisition of DeciderAI,
resulting in the formation of a new merged entity, HotShotAI. Although the acquisition was
finalized, the integration of the two former companies’ systems, processes and cultures is
still a work in progress.
O’Sullivan’s expectation is for DeciderAI’s technology to be integrated in the HotShot
prototype for the demonstration. Per the terms of the acquisition, the two companies had
to be firewalled until the acquisition completed. On September 1, 2024, four months prior
to the planned acquisition completion date, HotShot subcontracted DeciderAI to develop
the AI based logistic and resource deployment model for their prototype solution. Now the
two teams must integrate their solutions into a joint HotShotAI prototype.
6
Joint Prototype Solution
O’Sullivan, CEO of the new joint company HotShotAI, tasked Miller and Garcia to review
the solutions of both teams to develop a joint product for the NIFC by May 30, 2025.
With the acquisition complete, Garcia sees the entire scope of the project and feels a
fundamental change is needed to enable the joint architecture to be more open to future
technologies. Due to the existing relationship and trust between O’Sullivan and Miller,
Garcia must present a strong case to O’Sullivan, substantiating his ideas with evidence
and aligning them with project objectives, to counter potential bias towards Miller's
expertise.
Miller feels human input at every critical decision-making step is the only way to avoid
some of the pitfalls of AI, and she does not want bad data and biases impacting life and
property. She assures management she can meet the NIFC requirements with minimal AI
using advances in weather technologies, drones, sensors, weather balloons, and other
more established methods. These systems can be maximized while leaving final decisions
up to humans who have experience and understand the intricacies of the situations which
might be faced in a fire. She feels human decision making or intervention is required in the
solution to maintain a safe and compliant process minimizing impacts to life, property and
resources during deployment.
Miller and Garcia have clashed since the beginning of the acquisition process on how best
to design and lead the engineering organization in the new company with both competing
for the Chief Engineer position in the combined company. Dr. Brian Pegg, a senior
engineer, previously from HotShot who reported to O’Sullivan, believes Miller provided
incomplete data to Garcia prior to the acquisition close which could negatively affect
Garcia’s solution. Garcia picked up on this without HotShot’s knowledge, and he feels
Miller is deliberately trying to discredit his solution.
The outcome of this project will be a deciding factor in their respective career paths. While
working on the FIREINT architecture, Miller and Garcia struggle to agree on a common
solution. Their clashes have become more frequent and more public with yelling and name
calling. Miller thinks every decision is critical whereas Garcia feels there are only a few
decisions that cannot be made by AI. Some feel they are trying to sabotage the other’s
solution.
Two engineers, Dan Wells, previously HotShot, and Erin Kay, previously DeciderAI, are
getting tired of not being heard in meetings and the bickering between Miller and Garcia.
Wells has heard Miller say she has information she is not giving Garcia and lets Kay know.
Wells and Kay weren’t comfortable, on their own, to approach management with their
concerns but together they felt they need to report their observations. Wells and Kay don’t
know how to raise concerns at HotShotAI. They are unsure what normal debate between
senior colleagues looks like and when it becomes unacceptable. With the company being
new, they don’t know who to trust or how to speak up. When the debate between Miller
and Garcia turns into yelling, they talked to their managers, but the behavior continued
and the team continues to rehash the same old arguments while no progress is made
towards project objectives. It is now to the point where no action is being taken to address
the conflict between Miller and Garcia, and Wells and Kay are now questioning whether
HotShotAI is the place for them and heard rumors others are thinking the same thing.
7
Miller and Garcia continue to fight over whether human intervention is always needed. For
example, Miller feels a manned reconnaissance flight is always needed prior to deploying
fixed wing assets, but Garcia feels drones and AI solutions can be used to determine if safe
conditions exist to deploy these assets.
Miller and Garcia also cannot agree on whether the level of bias introduced by AI and its
impact outweighs the drawbacks of limited information considered in human decision
making. Miller feels there is bias in the AI which will lead to more affluent areas prioritized
over low-income housing areas. The AI algorithms use government census data and hence
Garcia feels there is no bias. Garcia's AI algorithm may have unintentionally reflected
socioeconomic biases, potentially impacting the equitable distribution of resources in wildfire
response scenarios, underscoring the need for algorithmic auditing and testing to ensure
fairness and accuracy.
Response to FIREINT Proposal
Since Miller and Garcia are not making much progress with the integration, O’Sullivan called
a meeting during the week of February 24, 2025, for their teams to resolve the issues and
determine the final solution. During this meeting the teams must present defendable
rationale on a combined solution and agree how to integrate their respective solutions into
the best single solution integrating AI, technology, and human intelligence and decision
making as well as the costs and timelines to complete the prototype to present to the
HotShotAI senior management team in 14 days. Loss of this contract could have impacts on
the company’s future. The goal of the meeting is to determine the most effective approach
for integrating the existing capabilities of both companies. (See Figure 2 for project timeline.)
During this meeting, the teams must agree on where and when human intervention/decision
making is needed in the solution, how their solution will decide which resources to deploy
where and how to prioritize the risks including threat to life (including firefighters), and threat
to property. They must also show how the joint solution will meet all NIFC requirements.
8
Figure 1: Map of Area to be Used for Prototype
Area of eastern Orange County, CA. The area is impacted by the Santa Ana winds where the prevailing winds blow from east to west
over the mountains towards the Pacific coast.
9
Figure 2: Project Timeline
– Slide Number 1
– Slide Number 2
– Slide Number 3
– Slide Number 4
– Slide Number 5
– Slide Number 6
– Slide Number 7
– Figure 1: Map of Area to be Used for Prototype
– Figure 2: Project Timeline
– Slide Number 10
– Slide Number 11
– Slide Number 12
**## Case.pdf**
2
Please Note: The fictional case provided herein was
developed over the past seven months, before the current
California fires. Lockheed Martin is deeply committed to
supporting the relief and recovery efforts of those affected by
the fires. We have donated to support immediate relief and
long-term recovery of communities impacted by the Los
Angeles wildfires. You can learn more about our efforts here.
We understand that the theme of firefighting intelligence is
particularly relevant but also sensitive given the current
situation. We want to acknowledge that Lockheed Martin
products are being used to assist in the firefighting efforts in
Los Angeles. While our technology is designed to support
these efforts, we do not want to draw attention away from
the very real challenges and tragedies that are unfolding in
California. Our intention is to provide a thought-provoking
and educational experience for all participants, and we hope
that this case will serve as a valuable learning tool.
[https://www.lockheedmartin.com/en-us/news/statements-speeches/2025/los-angeles-wildfires-relief-and-recovery.html](https://www.lockheedmartin.com/en-us/news/statements-speeches/2025/los-angeles-wildfires-relief-and-recovery.html)
3
2025 Ethics in Engineering Case
Background
In 2023, there were over 56,000 wildfires across the US burning 2.6 million acres of land
and destroying over 4,300 structures including around 3,000 homes. The total cost of
these wildfires is estimated to be between $394 to $893 billion each year.
Analyzing wildfire intelligence information and deploying firefighting resources is a time-
consuming process that requires multiple manual decisions. This is a complex process
and due to time and resource constraints, involving humans in every decision is inefficient.
Because of the changing nature of wildfires, the current process is expensive and prone to
human error. The use of advanced technologies, including artificial intelligence (AI),
machine learning (ML), and other automated methods, could significantly reduce data
processing times and improve data accuracy. This would enable incident commanders,
division supervisors, and hand crew bosses to make critical decisions more rapidly,
thereby reducing risk to life, property, structures, and land.
To modernize the approach, reduce cost, and better serve those in high fire risk areas, the
National Interagency Fire Center (NIFC) is assessing new technologies to better manage
resources during wildfires. They are developing a firefighting intelligence prototype
solution called FIREINT (Firefighting Intelligence), which will be used to assess fire
conditions, analyze the environment and location of the fire, and determine the most
efficient use of resources, including human intervention in verification and decision
making, to combat the fire to best protect the surrounding area. The development of
FIREINT is expected to be completed by May 2025, with plans for the operational system
to be deployed prior to the peak of the 2026 wildfire season.
The objectives of FIREINT are to:
• maximize the use of technology while ensuring human involvement in critical decisions.
• assist incident commanders in prioritizing the deployment of scarce resources while
providing maximum protection to emergency responders.
• enable developers to easily integrate and swap out various AI algorithms, data sources,
and interfaces, allowing for rapid customization and adaptation to changing wildfire
response scenarios.
The Capture
The NIFC has contracted three companies to develop a prototype FIREINT solution to be
demonstrated to them in May 2025.
The NIFC has specified the requirements below within which the FIREINT demo shall
operate. A representative map highlighting key features associated with the demo is
included in Figure 1.
4
FIREINT shall:
1. be capable of assessing wildfires in an area of up to 200,000 acres
(approximately 300 square miles) with various terrains and population zones
found in the western and southwestern US to inform incident commanders
where to deploy resources.
2. notify when human decision making is required to minimize risk to human life
and resources.
3. base fire suppression solution on deployment of available ground based and
aerial resources. This consists of 23 air tankers, 11 transport helicopters, and 14
tactical coordination aircraft for fire reconnaissance. Other types of resources
and technology to support traditional firefighting resources can be part of the
solution.
4. provide near real-time information to incident commanders and emergency
responders.
5. explain the decision-making processes for when to use different levels of
technology and human interaction while taking into account the following
considerations.
a. Protection and safety of firefighters;
b. Prioritization of human lives versus property;
c. Protection of critical infrastructure needed to support homes and life;
d. Evaluation of environmental impact from both the fire and suppressant;
e. Optimization/preservation of fire suppression asset utilization; and
f. Prioritization of monetary and historic value impact.
Companies will be contracted up to a value of $1M for the prototype solution which
includes the demonstration. Following the demonstration of each company’s
solution, the NIFC will select the winning solution and award a follow-on contract
valued at $10M.
HotShot Response Inc. (HotShot)
HotShot is one of the three companies selected by the NIFC to develop the
prototype FIREINT solution by May 2025. HotShot has a long history of providing
exceptional support to the NIFC. They have a record of consistently achieving the
mission to significantly reduce casualties during high risk and constantly changing
fire and weather situations via their logistics and firefighting asset deployment
systems. This mission is deeply embedded into the culture of the company. The
existing architecture's limited openness and finite scalability hinder timely upgrades,
underscoring the need for a modular, open-source system that can seamlessly
integrate with diverse components and adapt to evolving wildfire response
requirements.
5
HotShot’s president, Kevin O’Sullivan, realized months before this opportunity came up,
they needed to modernize their current product line and develop discriminating capabilities
to remain a leading partner for the NIFC. He felt they had internal expertise to advance
automation, but they needed external resources to incorporate AI capabilities.
HotShot’s chief engineer, Kimberly Miller, is not in total agreement with needing AI to
advance their solution and continues to fight against the need for using it. They have used
their current strategy for the last 25 years with great success without needing outside help.
O’Sullivan respects Miller’s viewpoint but sees the future technology landscape changing
and recognizes the need to integrate automation and AI into their current solutions. As a
result, O’Sullivan entered negotiations to acquire DeciderAI.
DeciderAI
DeciderAI is an innovative company with a proven and fielded AI resource deployment and
logistics solution that addresses hurricane emergency management, but they lack the
resources to expand into adjacent markets. DeciderAI leverages a suite of key enabling
technologies to optimize the deployment and allocation of resources in complex, dynamic
environments. At the core of their platform lies a robust combination of AI algorithms,
which enable real-time predictive analytics, pattern recognition, and decision-making. Their
proprietary graph database management system provides a scalable and flexible data
structure for modeling and analyzing complex networks, allowing them to identify resource
limitations and predictive schedule conflicts.
DeciderAI’s Chief Engineer, Hunter Garcia, has been a pioneer in the field of AI for almost
20 years and, based on what he knows, he can develop a system with little to no human
intervention. He feels AI is smarter than any human and can take all bias out of the
decision-making process. Garcia doesn’t have firefighting intelligence experience but can
easily see how existing AI processes could be applied. In contrast to HotShot’s chief
engineer, Garcia feels the strategy and solutions should always be evolving to incorporate
the latest technologies. Those who stagnate get left behind.
HotShotAI – Acquisition of DeciderAI by HotShot
On December 31, 2024, HotShot successfully completed its acquisition of DeciderAI,
resulting in the formation of a new merged entity, HotShotAI. Although the acquisition was
finalized, the integration of the two former companies’ systems, processes and cultures is
still a work in progress.
O’Sullivan’s expectation is for DeciderAI’s technology to be integrated in the HotShot
prototype for the demonstration. Per the terms of the acquisition, the two companies had
to be firewalled until the acquisition completed. On September 1, 2024, four months prior
to the planned acquisition completion date, HotShot subcontracted DeciderAI to develop
the AI based logistic and resource deployment model for their prototype solution. Now the
two teams must integrate their solutions into a joint HotShotAI prototype.
6
Joint Prototype Solution
O’Sullivan, CEO of the new joint company HotShotAI, tasked Miller and Garcia to review
the solutions of both teams to develop a joint product for the NIFC by May 30, 2025.
With the acquisition complete, Garcia sees the entire scope of the project and feels a
fundamental change is needed to enable the joint architecture to be more open to future
technologies. Due to the existing relationship and trust between O’Sullivan and Miller,
Garcia must present a strong case to O’Sullivan, substantiating his ideas with evidence
and aligning them with project objectives, to counter potential bias towards Miller's
expertise.
Miller feels human input at every critical decision-making step is the only way to avoid
some of the pitfalls of AI, and she does not want bad data and biases impacting life and
property. She assures management she can meet the NIFC requirements with minimal AI
using advances in weather technologies, drones, sensors, weather balloons, and other
more established methods. These systems can be maximized while leaving final decisions
up to humans who have experience and understand the intricacies of the situations which
might be faced in a fire. She feels human decision making or intervention is required in the
solution to maintain a safe and compliant process minimizing impacts to life, property and
resources during deployment.
Miller and Garcia have clashed since the beginning of the acquisition process on how best
to design and lead the engineering organization in the new company with both competing
for the Chief Engineer position in the combined company. Dr. Brian Pegg, a senior
engineer, previously from HotShot who reported to O’Sullivan, believes Miller provided
incomplete data to Garcia prior to the acquisition close which could negatively affect
Garcia’s solution. Garcia picked up on this without HotShot’s knowledge, and he feels
Miller is deliberately trying to discredit his solution.
The outcome of this project will be a deciding factor in their respective career paths. While
working on the FIREINT architecture, Miller and Garcia struggle to agree on a common
solution. Their clashes have become more frequent and more public with yelling and name
calling. Miller thinks every decision is critical whereas Garcia feels there are only a few
decisions that cannot be made by AI. Some feel they are trying to sabotage the other’s
solution.
Two engineers, Dan Wells, previously HotShot, and Erin Kay, previously DeciderAI, are
getting tired of not being heard in meetings and the bickering between Miller and Garcia.
Wells has heard Miller say she has information she is not giving Garcia and lets Kay know.
Wells and Kay weren’t comfortable, on their own, to approach management with their
concerns but together they felt they need to report their observations. Wells and Kay don’t
know how to raise concerns at HotShotAI. They are unsure what normal debate between
senior colleagues looks like and when it becomes unacceptable. With the company being
new, they don’t know who to trust or how to speak up. When the debate between Miller
and Garcia turns into yelling, they talked to their managers, but the behavior continued
and the team continues to rehash the same old arguments while no progress is made
towards project objectives. It is now to the point where no action is being taken to address
the conflict between Miller and Garcia, and Wells and Kay are now questioning whether
HotShotAI is the place for them and heard rumors others are thinking the same thing.
7
Miller and Garcia continue to fight over whether human intervention is always needed. For
example, Miller feels a manned reconnaissance flight is always needed prior to deploying
fixed wing assets, but Garcia feels drones and AI solutions can be used to determine if safe
conditions exist to deploy these assets.
Miller and Garcia also cannot agree on whether the level of bias introduced by AI and its
impact outweighs the drawbacks of limited information considered in human decision
making. Miller feels there is bias in the AI which will lead to more affluent areas prioritized
over low-income housing areas. The AI algorithms use government census data and hence
Garcia feels there is no bias. Garcia's AI algorithm may have unintentionally reflected
socioeconomic biases, potentially impacting the equitable distribution of resources in wildfire
response scenarios, underscoring the need for algorithmic auditing and testing to ensure
fairness and accuracy.
Response to FIREINT Proposal
Since Miller and Garcia are not making much progress with the integration, O’Sullivan called
a meeting during the week of February 24, 2025, for their teams to resolve the issues and
determine the final solution. During this meeting the teams must present defendable
rationale on a combined solution and agree how to integrate their respective solutions into
the best single solution integrating AI, technology, and human intelligence and decision
making as well as the costs and timelines to complete the prototype to present to the
HotShotAI senior management team in 14 days. Loss of this contract could have impacts on
the company’s future. The goal of the meeting is to determine the most effective approach
for integrating the existing capabilities of both companies. (See Figure 2 for project timeline.)
During this meeting, the teams must agree on where and when human intervention/decision
making is needed in the solution, how their solution will decide which resources to deploy
where and how to prioritize the risks including threat to life (including firefighters), and threat
to property. They must also show how the joint solution will meet all NIFC requirements.
8
Figure 1: Map of Area to be Used for Prototype
Area of eastern Orange County, CA. The area is impacted by the Santa Ana winds where the prevailing winds blow from east to west
over the mountains towards the Pacific coast.
9
Figure 2: Project Timeline
– Slide Number 1
– Slide Number 2
– Slide Number 3
– Slide Number 4
– Slide Number 5
– Slide Number 6
– Slide Number 7
– Figure 1: Map of Area to be Used for Prototype
– Figure 2: Project Timeline
– Slide Number 10
– Slide Number 11
– Slide Number 12