I think the user case can be further breakdown to "Blind people taking elevator alone" because otherwise people can offer help. We can assume a 5 second "no-action" time inside a elevator monitor, but confirm that the person inside is alone and has visual incapability. Then, the system can automatically switch over to full kinect (motion control) mode (then all the high techs kick in). It could further be improved if after 10 seconds still no command mode, the elevator will automatically park safely and open the door to prevent any urgent situation.
Thanks Diego for the wonderful session. I love your PM interviews and find it very helpful for my interview preparation. Here is my moonshot idea. Alexa in Health care. Details in your message.
I could think of a really simple solution. 1. If we have a blind/partial vision impaired person taking the elevator, then sort of a pre-decided floor key card to access the elevator and reach the floor should be good (considering the privacy issues). For visitors, a similar thing can be done from the reception itself. The scanner for the keycard will be at every level. If the person wishes to go to a different floor, there will be a pressure point system (tap system) which they can use from outside the elevator, located near the access key scanner. Till the 5th floor you need to tap the screen 5 times. Beyond that there will be interment of 5. Let's assume the person wants to go to the 7th floor. So the person has to long press for 2 seconds, which means you have added 5 floors and now tap 2 times. 2. How do you know which elevator has arrived and which is going up or down and if there is space in the elevator? Here the system will only trigger when the access key is used. The elevator system will evaluate the one with lesser crowd and send it to the floor the blind person is at.
maybe have a conveyor belt - like movable platform next to the elevator where the blind person (and possibly their assistant) can stand on and be transported into to the destination elevator door in an automated manner. This system can be used by people with other disabilities as well.
I would suggest 1) Camera system on top of the elevator, detect people with Stick, people with black glasses or If someone is standing in front of the elevator even the elevator is on the floor (to understand that this person is with special need). 2) there will be speaker system as well. 1) it prompts up coming elevators. Let it be 1-2-3 with prompting the direction of the elevator. 2) it prompts, you are in front of elevetor 1-2-3 as well as the direction towards, where are the buttons to call the elevator. 3) 1) once the elevator is on the floor and the elevator is desired for the person. There will be prompt on door opening. 2) once the person gets inside the elevator and other camera and speaker system to guide the user regarding, Buttons or just say the floor number (here we can use speech to text) 3) prompt of closing door, The coming floor, prompt of opening of door, and closing.
Interviewee did not: 1. Mention how costs will factor into his solution despite menitoning it in the beginning 2. Define the user properly. There could be children that are blind, adults that are blind, elderly people that are blind. They could also have other disabilities associated with them. They also have different characteristics. Example children and elderly are probably almost always going to be accompanied 2b) Prioritze the user segment and reason why he chose to prioritize it 3. Define painpoints for his (un)chosen user segment 4. Prioritse unsaid painpoints 5. Detail how elevators currently work 6. Tries to solve every part of the user journey instead of prioritizing (would have said pain points but he didn't list any) 7. Give enough time to evaluate tradeoffs and spent too much time in the weeds of the solutions 8. Distinguish how his solution would differ from a domestic elevator
Listing down the user PAIN POINT was important which was missed and the interviewee moved to the solution. Its important to prioritize the pain point and make the solutions.
Absolutely! Having a solid understanding on the pain points and what we should solve for is critical for these problems. Thanks for commenting, Sachin!
My moonshot idea would be designing a mobile application with proximity capabilities and adapters on elevators which is integrated to elevator system rather than designing a complete new elevator or investing in camera systems, speakers etc. Hence a conventional elevator could be upgraded to blind-safe elevator level with less investment and with the integration of calendars of people, it would also possible to call elevator autonomously with no additional effort when they arrive into entrance floor. Blind people just wear their earphones and leave rest to the app and communicates with their voice.
Btw, the work you do in here Diego, it is extremely invaluable!!! You touch people's life from the countries even you may not able to show on a map, but they appreciate your efforts with full of their heart :)
Thank you so much, Tarik! And thanks for sharing these ideas!! I like it! My question is: what would you need to do to make sure this work with conventional elevators if you won’t add/install anything new?
@@PMDiegoGranados Hey Diego! If we are not allowed to integrate anything as a constraint, the only way is keeping our mobile at chest level with open camera to process elevator doors for us and talk to our earphones. Mobile app (alternatively oculus), empowered with tiny ml computer vision algo inside to; 1- Tell user where are the elevator doors and explain the path to call button through earphones 2- Monitor arriving elevators and direct the user next to arriving door 3- Direct the user into elevator while processing elevator control panel inside 4- Take floor request by speech recognition 5- Direct the users finger to correct button by processing numbers on buttons or panel 6- Inform when arrived to floor and direct to outside. Actually in my answer with "adapters on elevators which is integrated to elevator system" I had mentioned to design an integrated box which will communicate with our mobile app as adapter to make our solution elevator vendor-agnostic. Voice messages to be relayed to elevator system would be possible and proximity sensors would be possible with control plate. A small integrated box with proximity sensors and elevator controllers for top 10 elevator systems maybe Schindler, Hyundai, Otis, Hitachi, Kone etc... These boxes can also be integrated to their elevator systems by elevator vendors as well in scope of business partnership.
It sure is a very complex question. I love the work and questions on this channel. I have made a practise that I read the title of the video. Brainstorm and pen down my ideas and framework then come back and watch the video along with my notes. I have learnt a ton of things by doing this
This was a great question & I like how the interviewee was confident throughout. Thank you! Building onto his features/solutions, I have 2 suggestions. Would love to know your thoughts: 1. Scenario where more than one elevator comes at the same time - Instead of just a sound, we can have a mechanism where the elevators announce (something like Left elevator is here, Middle elevator is here..) & have the elevators be open for a few more seconds than they usually would 2. I liked the idea of having buttons on the elevator rails - When the button is pressed, we can have a voice-controlled smart speaker with clear instructions. The speaker could ask "Which floor, if you're not sure, say not sure". If they mention a floor, then its captured. If not, the speaker asks where they are going and suggests a floor. In addition to that, it can end by saying something like "Please let me know other floor options if you want multiple selections?" and this can be used by sighted and visually impaired people alike. This speaker can also announce when a floor selected through voice control is reached.
Thanks for your comment Vidhatri - Follow up questions for you 😉 1) What if there's too much noice in the lobby? How would you solve for that? 2) What if 5 people enter at the same time, how do you know whom the Elevator should listen to?
@@PMDiegoGranados 1. Too much noise is very much possible, Cortana or Alexa would fail in this case, the reply would be asking the user to repeat the input. It's an assumption that blind people hear more carefully and use the system as per instructions. 2. If the NLP heard different unique options like floor 2, 4,6, lift would stopping at all floors it heard and obviously says the position very clearly so the people can act accordingly
I am not a product guy but my curiosity is why the elevators should be customised in a building. Why not carry a device which visualises the place and informs the person. Kind of Google map where the person is instructed by an audio device in terms of route and which directs the person towards an open elevator through sensors and informs the opening and closing of doors. This device can then be input with the map of any building and hence give a person the route till the elevator.just like a g map, the software can give a audio based info to the blind person on location and height which will take care of which floor. Just discovered your channel and have subscribed immediately.Awesome.
This is very interesting @Soundaram, thank you for sharing and subscribing! Follow-up questions: 1) Specific for blind people - if the lobby, or the place, is too crowded and listening to the phone can be complicated. How would you help them through the app to get to the elevator? 🤔 2) How would you measure that this implementation was successful? in other words, how would you know that they are using your app and that it is actually something useful?
@@PMDiegoGranados well I would just copy the basics of gmap with some minor adjustments, like , before starting the journey, the blind person gives a voice based input of the address into the app which includes destination floor which would mean that he or she has to go from 1st or ground floor to the destination floor, I would copy the vibration part from gmap, except it would vibrate from and to, so ground floor vibration and destination floor vibration, maybe I will have a time set to vibrate for 10 seconds. In terms of measuring the usage, wouldn't we know if the app has been used and I will reward for buildings covered or distance travelled with say Braille books or some device etc. Actually I would look at this device to be used by blind and non blind both to find their way in building or large areas with multi building to reach an office or floor. Hey thanks for responding, when I saw your response I thought to myself maybe I am OK in imagining stuff😊
I love how organized and clear the answer is. However, I feel touch and audio are currently more feasible options than camera and AI detection which brings its own set of constraints. A button along the general hand level which can be accessible to wheelchair-bound, as well as blind people can summon an elevator that's going up/down, and another single button inside can assist with getting off at the right floors and emergencies. Auto-detection sensors in front of an elevator is also a great idea, but brings up a complication of elevator directions. Also, our user is sometimes accompanied by a service animal, which creates a few more edge cases that can be resolved by announcements and audio of door opening, closing, directions, whether elevator is at capacity etc. I'd like to know which approach is preferred - a more futuristic answer with a lot of tech or a simpler solution by design?
Thank you for asking Namratha! My recommendation is to say both as possible solutions - the futuristic approach & the simple one, but prioritize based on reality. For example, you can say that you will implement first the simple solution given its impact and smaller timeframe needed to develop, however, you will also start an R&D work stream for the futuristic scenario because it will make a difference in the next X years. Let me know if this helps
I would like to add on the feature, provide a touch pad where user can draw or write the floor number and a voice response to confirm the floor number, more helpful for all users
Thank you for asking Srishti! Here are a few metrics on User Experience: * Time on task * Task Success * Pageviews * Time Spent on Page * Session length * Task success rate: Number of correctly completed tasks / Total number of attempts. And these are just some examples, but if you search online for UX success metrics or user experience success metrics you can find plenty of options :)
Hi Diego, what do you think about a solution that is an app that can give instructions to the elevator? My assumption is that the blind user might have a mobile phone and likely have some assistance apps on it already. Having a "remote control" on your phone can actually solve the entire user problem. Also, this requires less efforts (development of an app, bluetooth connection to the elevator, sensor/control mechanism that can be wired to the elevator). My only concern is that 1) this solution sounds like avoiding the hard problem (software is easier than hardware) and 2) not really redesigning a physical elevator. What do you think? Appreciate it! p.s. Now that I'm into the latter part of the video, I think the candidate actually talked about the same solution :)
Ming, that's an interesting approach, thanks for sharing it! And yes, a mobile app is a nice solution to this problem... assuming the user frequently uses the elevator, what if it's a visitor? would you ask them to download the elevator while in the lobby? is there a better way to help those who are just visitors?
What is your moonshot idea? How would you design the elevator?
I think the user case can be further breakdown to "Blind people taking elevator alone" because otherwise people can offer help. We can assume a 5 second "no-action" time inside a elevator monitor, but confirm that the person inside is alone and has visual incapability. Then, the system can automatically switch over to full kinect (motion control) mode (then all the high techs kick in). It could further be improved if after 10 seconds still no command mode, the elevator will automatically park safely and open the door to prevent any urgent situation.
Thanks Diego for the wonderful session. I love your PM interviews and find it very helpful for my interview preparation. Here is my moonshot idea. Alexa in Health care. Details in your message.
I could think of a really simple solution. 1. If we have a blind/partial vision impaired person taking the elevator, then sort of a pre-decided floor key card to access the elevator and reach the floor should be good (considering the privacy issues). For visitors, a similar thing can be done from the reception itself. The scanner for the keycard will be at every level. If the person wishes to go to a different floor, there will be a pressure point system (tap system) which they can use from outside the elevator, located near the access key scanner. Till the 5th floor you need to tap the screen 5 times. Beyond that there will be interment of 5. Let's assume the person wants to go to the 7th floor. So the person has to long press for 2 seconds, which means you have added 5 floors and now tap 2 times.
2. How do you know which elevator has arrived and which is going up or down and if there is space in the elevator? Here the system will only trigger when the access key is used. The elevator system will evaluate the one with lesser crowd and send it to the floor the blind person is at.
maybe have a conveyor belt - like movable platform next to the elevator where the blind person (and possibly their assistant) can stand on and be transported into to the destination elevator door in an automated manner. This system can be used by people with other disabilities as well.
I would suggest
1) Camera system on top of the elevator, detect people with Stick, people with black glasses or If someone is standing in front of the elevator even the elevator is on the floor (to understand that this person is with special need).
2) there will be speaker system as well. 1) it prompts up coming elevators. Let it be 1-2-3 with prompting the direction of the elevator. 2) it prompts, you are in front of elevetor 1-2-3 as well as the direction towards, where are the buttons to call the elevator.
3)
1) once the elevator is on the floor and the elevator is desired for the person. There will be prompt on door opening.
2) once the person gets inside the elevator and other camera and speaker system to guide the user regarding, Buttons or just say the floor number (here we can use speech to text)
3) prompt of closing door, The coming floor, prompt of opening of door, and closing.
Interviewee did not:
1. Mention how costs will factor into his solution despite menitoning it in the beginning
2. Define the user properly. There could be children that are blind, adults that are blind, elderly people that are blind. They could also have other disabilities associated with them. They also have different characteristics. Example children and elderly are probably almost always going to be accompanied
2b) Prioritze the user segment and reason why he chose to prioritize it
3. Define painpoints for his (un)chosen user segment
4. Prioritse unsaid painpoints
5. Detail how elevators currently work
6. Tries to solve every part of the user journey instead of prioritizing (would have said pain points but he didn't list any)
7. Give enough time to evaluate tradeoffs and spent too much time in the weeds of the solutions
8. Distinguish how his solution would differ from a domestic elevator
This is a very insightful mock interview. Thanks a ton for sharing this.
Listing down the user PAIN POINT was important which was missed and the interviewee moved to the solution. Its important to prioritize the pain point and make the solutions.
Absolutely! Having a solid understanding on the pain points and what we should solve for is critical for these problems. Thanks for commenting, Sachin!
My moonshot idea would be designing a mobile application with proximity capabilities and adapters on elevators which is integrated to elevator system rather than designing a complete new elevator or investing in camera systems, speakers etc. Hence a conventional elevator could be upgraded to blind-safe elevator level with less investment and with the integration of calendars of people, it would also possible to call elevator autonomously with no additional effort when they arrive into entrance floor. Blind people just wear their earphones and leave rest to the app and communicates with their voice.
Btw, the work you do in here Diego, it is extremely invaluable!!!
You touch people's life from the countries even you may not able to show on a map, but they appreciate your efforts with full of their heart :)
Thank you so much, Tarik! And thanks for sharing these ideas!!
I like it! My question is: what would you need to do to make sure this work with conventional elevators if you won’t add/install anything new?
@@PMDiegoGranados Hey Diego!
If we are not allowed to integrate anything as a constraint, the only way is keeping our mobile at chest level with open camera to process elevator doors for us and talk to our earphones.
Mobile app (alternatively oculus), empowered with tiny ml computer vision algo inside to;
1- Tell user where are the elevator doors and explain the path to call button through earphones
2- Monitor arriving elevators and direct the user next to arriving door
3- Direct the user into elevator while processing elevator control panel inside
4- Take floor request by speech recognition
5- Direct the users finger to correct button by processing numbers on buttons or panel
6- Inform when arrived to floor and direct to outside.
Actually in my answer with "adapters on elevators which is integrated to elevator system" I had mentioned to design an integrated box which will communicate with our mobile app as adapter to make our solution elevator vendor-agnostic. Voice messages to be relayed to elevator system would be possible and proximity sensors would be possible with control plate. A small integrated box with proximity sensors and elevator controllers for top 10 elevator systems maybe Schindler, Hyundai, Otis, Hitachi, Kone etc... These boxes can also be integrated to their elevator systems by elevator vendors as well in scope of business partnership.
It sure is a very complex question. I love the work and questions on this channel. I have made a practise that I read the title of the video. Brainstorm and pen down my ideas and framework then come back and watch the video along with my notes. I have learnt a ton of things by doing this
Thanks for sharing!
This was a great question & I like how the interviewee was confident throughout. Thank you!
Building onto his features/solutions, I have 2 suggestions. Would love to know your thoughts:
1. Scenario where more than one elevator comes at the same time - Instead of just a sound, we can have a mechanism where the elevators announce (something like Left elevator is here, Middle elevator is here..) & have the elevators be open for a few more seconds than they usually would
2. I liked the idea of having buttons on the elevator rails - When the button is pressed, we can have a voice-controlled smart speaker with clear instructions. The speaker could ask "Which floor, if you're not sure, say not sure". If they mention a floor, then its captured. If not, the speaker asks where they are going and suggests a floor. In addition to that, it can end by saying something like "Please let me know other floor options if you want multiple selections?" and this can be used by sighted and visually impaired people alike. This speaker can also announce when a floor selected through voice control is reached.
Thanks for your comment Vidhatri - Follow up questions for you 😉
1) What if there's too much noice in the lobby? How would you solve for that?
2) What if 5 people enter at the same time, how do you know whom the Elevator should listen to?
@@PMDiegoGranados 1. Too much noise is very much possible, Cortana or Alexa would fail in this case, the reply would be asking the user to repeat the input. It's an assumption that blind people hear more carefully and use the system as per instructions. 2. If the NLP heard different unique options like floor 2, 4,6, lift would stopping at all floors it heard and obviously says the position very clearly so the people can act accordingly
Can't we just introduce a voice assistant like Alexa in the elevator? It is a universal approach with Braille enabled keys.
Informative mock interview
I am not a product guy but my curiosity is why the elevators should be customised in a building. Why not carry a device which visualises the place and informs the person.
Kind of Google map where the person is instructed by an audio device in terms of route and which directs the person towards an open elevator through sensors and informs the opening and closing of doors. This device can then be input with the map of any building and hence give a person the route till the elevator.just like a g map, the software can give a audio based info to the blind person on location and height which will take care of which floor. Just discovered your channel and have subscribed immediately.Awesome.
This is very interesting @Soundaram, thank you for sharing and subscribing! Follow-up questions:
1) Specific for blind people - if the lobby, or the place, is too crowded and listening to the phone can be complicated. How would you help them through the app to get to the elevator? 🤔
2) How would you measure that this implementation was successful? in other words, how would you know that they are using your app and that it is actually something useful?
@@PMDiegoGranados well I would just copy the basics of gmap with some minor adjustments, like , before starting the journey, the blind person gives a voice based input of the address into the app which includes destination floor which would mean that he or she has to go from 1st or ground floor to the destination floor, I would copy the vibration part from gmap, except it would vibrate from and to, so ground floor vibration and destination floor vibration, maybe I will have a time set to vibrate for 10 seconds.
In terms of measuring the usage, wouldn't we know if the app has been used and I will reward for buildings covered or distance travelled with say Braille books or some device etc.
Actually I would look at this device to be used by blind and non blind both to find their way in building or large areas with multi building to reach an office or floor.
Hey thanks for responding, when I saw your response I thought to myself maybe I am OK in imagining stuff😊
I love how organized and clear the answer is. However, I feel touch and audio are currently more feasible options than camera and AI detection which brings its own set of constraints. A button along the general hand level which can be accessible to wheelchair-bound, as well as blind people can summon an elevator that's going up/down, and another single button inside can assist with getting off at the right floors and emergencies.
Auto-detection sensors in front of an elevator is also a great idea, but brings up a complication of elevator directions. Also, our user is sometimes accompanied by a service animal, which creates a few more edge cases that can be resolved by announcements and audio of door opening, closing, directions, whether elevator is at capacity etc.
I'd like to know which approach is preferred - a more futuristic answer with a lot of tech or a simpler solution by design?
Thank you for asking Namratha!
My recommendation is to say both as possible solutions - the futuristic approach & the simple one, but prioritize based on reality. For example, you can say that you will implement first the simple solution given its impact and smaller timeframe needed to develop, however, you will also start an R&D work stream for the futuristic scenario because it will make a difference in the next X years.
Let me know if this helps
@@PMDiegoGranados that does help! Thanks a lot :)
I would like to add on the feature, provide a touch pad where user can draw or write the floor number and a voice response to confirm the floor number, more helpful for all users
We could have NLP system. Braille buttons for calling the lift and floors works.
Thanks for sharing this! Follow-up question: How would you use the NLP system? what is the user experience around it?
@@PMDiegoGranados the lift can be called and operated with voice commands.
That was one another great interview and amazing feedback. What are you thoughts when it comes to metrics when one is designing for experience?
Thank you for asking Srishti! Here are a few metrics on User Experience:
* Time on task
* Task Success
* Pageviews
* Time Spent on Page
* Session length
* Task success rate: Number of correctly completed tasks / Total number of attempts.
And these are just some examples, but if you search online for UX success metrics or user experience success metrics you can find plenty of options :)
Hi Diego, what do you think about a solution that is an app that can give instructions to the elevator? My assumption is that the blind user might have a mobile phone and likely have some assistance apps on it already. Having a "remote control" on your phone can actually solve the entire user problem. Also, this requires less efforts (development of an app, bluetooth connection to the elevator, sensor/control mechanism that can be wired to the elevator).
My only concern is that 1) this solution sounds like avoiding the hard problem (software is easier than hardware) and 2) not really redesigning a physical elevator. What do you think? Appreciate it!
p.s.
Now that I'm into the latter part of the video, I think the candidate actually talked about the same solution :)
Ming, that's an interesting approach, thanks for sharing it!
And yes, a mobile app is a nice solution to this problem... assuming the user frequently uses the elevator, what if it's a visitor? would you ask them to download the elevator while in the lobby? is there a better way to help those who are just visitors?
How would the blind person navigate or use the mobile app?
Here you go 😉 hope this helps:
th-cam.com/video/Hx4ivoI_GmM/w-d-xo.html