Redesigning Lumeto’s crisis response simulations to reduce error by 70%
Date: Jan 2022 - June 2022
Role: UX Designer, UX Researcher
How it works
1
Immersive VR Training
Learners roleplay to de-escalate mental health crisis scenarios
1
Immersive VR Training
Learners roleplay to de-escalate mental health crisis scenarios
2
6 Unique Scenarios
2
6 Unique Scenarios
3
Interactive Person in Crisis
Trainers control the actions and responses of the person in crisis to test their learners from our desktop app.
3
Interactive Person in Crisis
Trainers control the actions and responses of the person in crisis to test their learners from our desktop app.
Impact
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
My role
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
Research
Identify and solve critical usability issues
When I joined the product team as the sole designer, we were facing several critical issues. The team was six months away from its contractual delivery date but the product was still in its early stages. Initial testing indicated that the product wasn’t effective and was frustrating to use. We needed to figure out what wasn’t working and build a solution quickly, or risk losing our contract with the Ministry of the Solicitor General.
Identify and solve critical usability issues
When I joined the product team as the sole designer, we were facing several critical issues. The team was six months away from its contractual delivery date but the product was still in its early stages. Initial testing indicated that the product wasn’t effective and was frustrating to use. We needed to figure out what wasn’t working and build a solution quickly, or risk losing our contract with the Ministry of the Solicitor General.
1
Low scenario completion rate
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High degree of user error
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
Insights
Live user testing at our studio
Live user testing at our studio
Research data collected and analyzed in dovetail
Research data collected and analyzed in dovetail
Usability test footage
37 hours
Participants
43 users
Exploration
1
Low Findability
Trainers struggled to find responses within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
1
Low Findability
Trainers struggled to find responses within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High cognitive load
Trainers were unable to observe and evaluate learners while being required to control the person in crisis.
3
Missing functionality
Trainers consistently lacked the actions or functionality required to respond appropriately to their Learners, resulting in immersion breaking errors.
4
Low immersion
Trainer error or delayed response time was causing significant friction within the experience, breaking immersion for the learner and no longer effectively simulating a high-stakes crisis situation.
4
Learner agency is critical
Due to the open-ended nature of live roleplay from the Learners in VR, our desktop Trainer interface needed to provide a diverse array of actions and responses for the Person in Crisis. Based on the results of user testing, the way scenario content had been planned was insufficient for creating the level of conversational nuance required for an effective training product.
It became evident to me that our scenario’s narrative framework needed to be re-designed to support a more non-linear experience.
Impact
1
Analyzed actor-led training
Partnered with SMEs to gain a better understanding of the ways our Persons in Crisis should respond to Police actions.
2
Re-designed the User Journey and Dialogue Branching
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
3
Validated with the team
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
Final design
3
Dialogue sorted by context
Actions and dialogue were grouped according to their narrative context. The Trainer would select which contextual group to open from the main action list, and then be presented with all relevant actions or lines of dialogue related to that topic.
We observed that in the majority of cases the next required option would be within the same contextual group, drastically increasing findability and reducing trainer response time.
2
Unnecessary dialogue hidden
We observed a number of actions or responses across multiple scenarios that are only relevant at the beginning of the scenario, or at the very end. In order to reduce error and cognitive load, we created a further way to filter actions and responses according to narrative progression in the scenario.
3
Greater immersion with variable dialogue
In order to prevent the Person in Crisis from sounding robotic and breaking Learner Immersion, we enabled frequently used actions to trigger a series of distinct lines of dialogue with the same narrative implication. This allowed the Trainer to create a much richer and immersive conversation without adding additional buttons to manage.
4
Responses reflect escalation
Previously, dialogue options were locked behind levels of escalation. Testing revealed that Trainers were treating these escalation levels as containers for content, rather than narratively significant tools to respond to Learner actions.
To remedy this we exposed all possible actions to the Trainer at every level of escalation, with actions triggering a unique voice-line reflective of the character’s emotional state.
Post launch
1
Created a process for onboarding new Police Services
After the successful launch of our product, I leveraged insights gained from user research to develop a comprehensive onboarding process tailored for new Police Services. Training events were held at our studio for key members from each Police Service, familiarizing them with our product as well as the new training curriculum it was built upon. This resulted in a significantly faster onboarding process, allowing users to begin training with our product almost immerdiately.
Usability test footage
129 hours
Participants
112 users
2
Pilot Study on VR Training Efficacy
Working alongside Wilfred Laurier and Ryerson Universities, we conducted a pilot study to measure the efficacy of our simulation platform against traditional actor-led training.
2
Pilot Study on VR Training Efficacy
Working alongside Wilfred Laurier and Ryerson Universities, we conducted a pilot study to measure the efficacy of our simulation platform against traditional actor-led training.
3
Onsite deployment
Partnered with SMEs to gain a better understanding of the ways our Persons in Crisis should respond to Police actions.
Next Steps
Convert deployment insights into product improvements
Initial user feedback indicated critical problems with our approach to the product severely undermining its effectiveness as a training tool
You might also like...
Redesigning Lumeto’s crisis response simulations to reduce error by 70%
Date: Jan 2022 - June 2022
Role: UX Designer, UX Researcher
How it works
1
Immersive VR Training
Learners in VR roleplay to de-escalate mental health crisis scenarios.
2
6 Unique Scenarios
3
Interactive Person in Crisis
Trainers control the actions and responses of the person in crisis to test their learners from our desktop app.
Impact
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
My role
Identify and solve critical usability issues
When I joined the product team as the sole designer, we were facing several critical issues. The team was six months away from its contractual delivery date but the product was still in its early stages. Initial testing indicated that the product wasn’t effective and was frustrating to use. We needed to figure out what wasn’t working and build a solution quickly, or risk losing our contract with the Ministry of the Solicitor General.
1
Low scenario completion rate
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High degree of user error
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
Research
Live user testing at our studio
Usability test footage
37 hours
Participants
43 users
Research data collected and analyzed in dovetail
Insights
1
Low Findability
Trainers struggled to find responses within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High Cognitive Load
Trainers were unable to observe and evaluate learners while being required to control the person in crisis.
3
Missing functions
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
4
Low immersion
Trainer error or delayed response time was causing significant friction within the experience, breaking immersion for the learner and no longer effectively simulating a high-stakes crisis situation.
4
Learner agency is critical
Due to the open-ended nature of live roleplay from the Learners in VR, our desktop Trainer interface needed to provide a diverse array of actions and responses for the Person in Crisis. Based on the results of user testing, the way scenario content had been planned was insufficient for creating the level of conversational nuance required for an effective training product.
It became evident to me that our scenario’s narrative framework needed to be re-designed to support a more non-linear experience.
Exploration
1
Analyzed actor-led training
Partnered with SMEs to gain a better understanding of the ways our Persons in Crisis should respond to Police actions.
2
Re-designed the User Journey and Dialogue Branching
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
3
Validated with the team
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
Final design
1
Dialogue sorted by context
Actions and dialogue were grouped according to their narrative context. The Trainer would select which contextual group to open from the main action list, and then be presented with all relevant actions or lines of dialogue related to that topic.
We observed that in the majority of cases the next required option would be within the same contextual group, drastically increasing findability and reducing trainer response time.
2
Unnecessary dialogue hidden
We observed a number of actions or responses across multiple scenarios that are only relevant at the beginning of the scenario, or at the very end. In order to reduce error and cognitive load, we created a further way to filter actions and responses according to narrative progression in the scenario.
3
Greater immersion with variable dialogue
In order to prevent the Person in Crisis from sounding robotic and breaking Learner Immersion, we enabled frequently used actions to trigger a series of distinct lines of dialogue with the same narrative implication. This allowed the Trainer to create a much richer and immersive conversation without adding additional buttons to manage.
4
Responses reflect escalation
Previously, dialogue options were locked behind levels of escalation. Testing revealed that Trainers were treating these escalation levels as containers for content, rather than narratively significant tools to respond to Learner actions.
To remedy this we exposed all possible actions to the Trainer at every level of escalation, with actions triggering a unique voice-line reflective of the character’s emotional state.
Post launch
1
Created a process for onboarding new Police Services
After the successful launch of our product, I leveraged insights gained from user research to develop a comprehensive onboarding process tailored for new Police Services. Training events were held at our studio for key members from each Police Service, familiarizing them with our product as well as the new training curriculum it was built upon. This resulted in a significantly faster onboarding process, allowing users to begin training with our product almost immerdiately.
Usability test footage
129 hours
Participants
112 users
2
Pilot Study on VR Training Efficacy
Working alongside Wilfred Laurier and Ryerson Universities, we conducted a pilot study to measure the efficacy of our simulation platform against traditional actor-led training.
Impact
Impact
My role
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
Research
Insights
Exploration
Final design
Redesigning Lumeto’s crisis response simulations to reduce error by 70%
Date: Jan 2022 - June 2022
Role: UX Designer, UX Researcher
1
Immersive VR Training
Learners roleplay to de-escalate mental health crisis scenarios
1
Immersive VR Training
Learners roleplay to de-escalate mental health crisis scenarios
2
6 Unique Scenarios
2
6 Unique Scenarios
3
Interactive Person in Crisis
Trainers control the actions and responses of the person in crisis to test their learners from our desktop app.
3
Interactive Person in Crisis
Trainers control the actions and responses of the person in crisis to test their learners from our desktop app.
Post launch
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
Identify and solve critical usability issues
When I joined the product team as the sole designer, we were facing several critical issues. The team was six months away from its contractual delivery date but the product was still in its early stages. Initial testing indicated that the product wasn’t effective and was frustrating to use. We needed to figure out what wasn’t working and build a solution quickly, or risk losing our contract with the Ministry of the Solicitor General.
Identify and solve critical usability issues
When I joined the product team as the sole designer, we were facing several critical issues. The team was six months away from its contractual delivery date but the product was still in its early stages. Initial testing indicated that the product wasn’t effective and was frustrating to use. We needed to figure out what wasn’t working and build a solution quickly, or risk losing our contract with the Ministry of the Solicitor General.
1
Low scenario completion rate
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High degree of user error
Trainers struggled to find responses
within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
Impact
Live user testing at our studio
Live user testing at our studio
Research data collected and analyzed in dovetail
Research data collected and analyzed in dovetail
Usability test footage
37 hours
Participants
43 users
1
Low Findability
Trainers struggled to find responses within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
1
Low Findability
Trainers struggled to find responses within the UI, spending upwards of a few seconds before selecting a response to Learner actions.
2
High Cognitive Load
Trainers were unable to observe and evaluate learners while being required to control the person in crisis.
3
Missing functionality
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
4
Low immersion
Trainer error or delayed response time was causing significant friction within the experience, breaking immersion for the learner and no longer effectively simulating a high-stakes crisis situation.
4
Low immersion
Trainer error or delayed response time was causing significant friction within the experience, breaking immersion for the learner and no longer effectively simulating a high-stakes crisis situation.
4
Learner agency is critical
Due to the open-ended nature of live roleplay from the Learners in VR, our desktop Trainer interface needed to provide a diverse array of actions and responses for the Person in Crisis. Based on the results of user testing, the way scenario content had been planned was insufficient for creating the level of conversational nuance required for an effective training product.
It became evident to me that our scenario’s narrative framework needed to be re-designed to support a more non-linear experience.
4
Learner agency is critical
Due to the open-ended nature of live roleplay from the Learners in VR, our desktop Trainer interface needed to provide a diverse array of actions and responses for the Person in Crisis. Based on the results of user testing, the way scenario content had been planned was insufficient for creating the level of conversational nuance required for an effective training product.
It became evident to me that our scenario’s narrative framework needed to be re-designed to support a more non-linear experience.
1
Analyzed actor-led training
Partnered with SMEs to gain a better understanding of the ways our Persons in Crisis should respond to Police actions.
2
Re-designed the User Journey and Dialogue Branching
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
3
Validated with the team
Trainers consistently lacked the actions or functionality required to maintain required interactions with the Learners.
3
Dialogue sorted by context
Actions and dialogue were grouped according to their narrative context. The Trainer would select which contextual group to open from the main action list, and then be presented with all relevant actions or lines of dialogue related to that topic.
We observed that in the majority of cases the next required option would be within the same contextual group, drastically increasing findability and reducing trainer response time.
2
Unnecessary dialogue hidden
We observed a number of actions or responses across multiple scenarios that are only relevant at the beginning of the scenario, or at the very end. In order to reduce error and cognitive load, we created a further way to filter actions and responses according to narrative progression in the scenario.
2
Unnecessary dialogue hidden
We observed a number of actions or responses across multiple scenarios that are only relevant at the beginning of the scenario, or at the very end. In order to reduce error and cognitive load, we created a further way to filter actions and responses according to narrative progression in the scenario.
3
Greater immersion with variable dialogue
In order to prevent the Person in Crisis from sounding robotic and breaking Learner Immersion, we enabled frequently used actions to trigger a series of distinct lines of dialogue with the same narrative implication. This allowed the Trainer to create a much richer and immersive conversation without adding additional buttons to manage.
4
Responses reflect escalation
Previously, dialogue options were locked behind levels of escalation. Testing revealed that Trainers were treating these escalation levels as containers for content, rather than narratively significant tools to respond to Learner actions.
To remedy this we exposed all possible actions to the Trainer at every level of escalation, with actions triggering a unique voice-line reflective of the character’s emotional state.
1
Created a process for onboarding new Police Services
After the successful launch of our product, I leveraged insights gained from user research to develop a comprehensive onboarding process tailored for new Police Services. Training events were held at our studio for key members from each Police Service, familiarizing them with our product as well as the new training curriculum it was built upon. This resulted in a significantly faster onboarding process, allowing users to begin training with our product almost immerdiately.
Usability test footage
129 hours
Participants
112 users
2
Pilot Study on VR Training Efficacy
Working alongside Wilfred Laurier and Ryerson Universities, we conducted a pilot study to measure the efficacy of our simulation platform against traditional actor-led training.
2
Pilot Study on VR Training Efficacy
Working alongside Wilfred Laurier and Ryerson Universities, we conducted a pilot study to measure the efficacy of our simulation platform against traditional actor-led training.
3
Onsite deploy-ment
Partnered with SMEs to gain a better understanding of the ways our Persons in Crisis should respond to Police actions.
1
18k+
Users
Upon launch InvolveXR was rolled out to active duty police officers across Ontario
2
2M+
Revenue
Successful launch of the product satisfied contractual requirements and secured additional funding for future development
3
-70%
User Error
Solving critical usability challenges drastically reduced user error and increased scenario success rate across the board
4
Version 1.0
Successfully launching the product on time cleared our first major milestone as a team and brought the product to market
You might also like...
Interested in working together? Lets chat!
Interested in working together? Lets chat!
Interested in working together? Lets chat!