Porter

About the Project

Healing on Hold:

Why do we believe nurses even need an assistant? Isn’t their job already to be an assistant to a doctor? While this may be true, hospitals are incredibly intricate & complex systems of information, supplies, and people, and it turns out the overload of work is hindering the ability for nurses to do medical work that can actually benefit doctors on a larger scale. Here are some statistics to show the gigantic implications of this issue.

A 2018 study shows that about 10% of a nurse’s shift time goes into non-nursing specific tasks, that can easily be delegated, such as fetching supplies, and performing clerical duties
To take it one step further, hunting and gathering is shown to waste 6.6% of their time during each shift.
Another study from Vizzia Tech & Georgia State University in 2024 suggests that nurses lose 60 minutes per shift searching for equipment, expanding this across the US alone, scales up to $14 Billion potential dollars of labor that can be used to service patients
Many studies show direct patient care only makes up 19% of their time, compared to documentation and charting, being their biggest time consumer, at about 27%. Although it is an important task, it is not direct, and seemingly is a waste of time
In a Canadian scope, The Cancer Advocacy Coalition of Canada reported that oncology nurses spent almost one-third of a nurse’s time on non-nursing tasks, some of which are classified as tasks outside the scope of a nurse, such as notifying patients of future appointments. All-in-all, we can see nurses across Canada and US easily spend 30% of their shift time on non-clinical work, which we established can cause a lot of issues, which is an issue which also extends to smaller medical centres and clinics. Think about the other implications of this problem as well, including mental health of nurses, inefficiency of reaching patients, and potential loss of attention to more serious nursing tasks

A Smarter Way to Support Nurses:

We have established that by cutting down on these repetitive tasks for nurses, we can start working towards faster turnaround times for hospitals, reduced costs, and overall improvement to the medical industry as a whole!

Our proposition is to create robot companions for nurses that would make their jobs easier, and we have ideas for potential features. Firstly, the robots would automate any tasks that are easy, repetitive, and relatively simple. Many nurses won’t have to worry about basic patient care needs, such as providing water, or pillows to customers, this would be activated through basic voice commands and existing hospital practices (such as service buttons)

If the patient can request supplies, so can the nurses and doctors, instead of medical workers grabbing things, robots can quickly and efficiently do so.
The robots can also follow nurses around and act as scribes, automating the process of documentation for nurses as well
Finally, the robots can also check vitals. All patients need their vitals (like blood pressure, or blood test results) taken every 2-4 hours, a robot can take a cuff and a small blood testing device around to automate this task for them as well. MULTIPLE MEDICAL WORKERS HAVE MENTIONED HAVING VITALS ON HAND TO BE A “GAME CHANGER”

The robot would be a simple line-following robot, with built in sensors and other appropriate technologies to complete the tasks:

Ultrasonic Sensors and Cameras for navigation, as well as line following IR sensors to keep them on track, as well as a smart backend system to allow the robot to navigate smartly and efficiently
A screen and microphone interface, to clearly communicate with patients and hospital staff, through a Speech to text Model. The Gemini API would allow us to power the communication and decision making aspect for the robot, ensuring it makes appropriate decisions
A built in application, to manage remote commands, hospital routes, user data that the car has access to, and other relevant settings related to PORTER

Potential (Extra) Use Cases and Users:

Hospitals: Would use on a large scale to automate repetitive and undesirable tasks for nurses, effectively removing any non-medical or patient related work, while saving billions in wasted labour costs
Smaller Clinics: May prefer to use the AI-powered voice assistant to assist with tasks involving patients, smaller clinics can accept more patients and find other uses from PORTER.
Nursing Homes: Would get a similar use out of these robots as hospitals, except these robots would be more focused on serving the elderly as opposed to being general assistants to the nurses

Market, Target Demographic, & Industry Traction:

Our Target Demographic: Primarily hospital owners, and medical technicians in charge of hospital technologies

Our Target User: Nurses, Doctors, Pediatricians, General Health Care Workers

The market share estimations below will be based on similar data collected from the introduction of the Real Time Location Systems in the healthcare sector. The RTLS allows staff to remember where things are more easily, by using bluetooth, or Wi-Fi technology to track where specific equipment was at a given time. Although it was only introduced in 2012, some form of RTLS has entered 25%-30% of hospitals in the USA (out of ≈6000). Our product will provide a similar benefit to nurses and doctors, but on a much larger scale, because it not only can work in conjunction with RTLS (to those places that have it), but also can use its more advanced assistance features. Knowing that RTLS which is significantly less advanced, but still reached so many people despite its’ price of multiple hundred thousand, there is no doubt that there is very high market potential for our product as well.

Total Addressable Market: A fully customizable robot experience, that would continue to help every corner of the health sector, and perhaps beyond it, whether a small clinic needs to buy a robot for basic needs, or by a large hospital with all sorts of requirements, a future, refined version of our robot can automate tasks for medical professionals all over the world, there is no medical facility that would not benefit from automation in one way or another, making this project and incredibly scalable, feasible, and applicable one. Large hospitals can buy multiple robots, perhaps 5-20 (depending on varying needs) Nursing homes or Small hospitals can draw benefit from even fewer robots, such as 2-5 Small clinics, or specialist offices can benefit from 1 or 2 robots Overall, with max market penetration, there are millions of medical institutions these robots can go to, many backed by the government, or large stakeholders, making the product incredibly viable on a global scale, especially in 1st world countries. Looking at North America alone, I think it is very possible for our product to at least match the 25%-30% of RTLS

Serviceable Addressable Market: Given a few years, with time to raise money, refine the product, and build more modules, the product clearly can benefit a lot of medical institutions. With growth and development in the semi-long term, our product can feasibly reach hundreds or of medical institutions spanning North America, with perhaps a little market penetration within Europe as well. This is a feasible starting point, especially given the generally higher amounts of money accessible to hospitals in these regions, as well as our initial product being made keeping American & Canadian nurses in mind. Clearly the robots will cater more to their culture and work environment, and will be justifiable purchase for these institutions Generally would be found more in larger hospitals as opposed to small clinics due to the early features prioritizing tasks in larger institutions Overall, the addressable market for the future is still quite large, and given the fact that they will solve real problems, the demand for them will be very real.

Serviceable Obtainable Market: In terms of market share in the near future, there is potential for quick growth, and willing participants. Given funding and a short-term time frame to fully build out this product, we can start getting them out to various hospitals in Canada & America fairly soon. Given the abundance of health care facilities in these countries, as well as the need for assistance to nurses. Within a short time frame, we can feasibly get 50-100 customers across the US & Canada. There is no doubt that this product will solve major problems, as well as being an ever-updating assistant, that is the next technological advancement for hospitals. This product can slowly start making its way into hospitals.

Business Model: The biggest pieces of value we provide to the hospitals are lower costs, and increased efficiency. Not only does out product reduce the wasted money going into “useless” labour as discussed earlier, but it also removes the burden from overworked nurses, improving their morale, their quality of service, and patient turnaround time. Our business thrives on being able to make the medical experience better for all involved in it, not only the patients, but the workers as well.

How we Make Money & Revenue Streams: It naturally would start off with selling robots, it would not be unjustified to say we can make a few thousand per sold robot, as well as paid upgrades or services, for those who may need more niche modules, as well as setup or maintenance costs, for first time installation, or upgrading of existing robots in a medical setting. With a mix of these three avenues of getting money, as well as warranty programs and other

Traction & Feedback from Real Medical Professionals: In terms of traction, we reached out to a couple different people in the medical industry throughout the world, mainly from Australia, Canada, & The USA. We received some overwhelmingly good feedback, as well as some advice, which gave us the idea to improve on our current idea. The majority of nurses, doctors, and medical technicians agreed that a robot assistant would be incredibly beneficial, especially in situations where medications, food, blankets, or supplies needed to be transported, however the medical professionals also gave us a perspective we hadn’t taken before. Which was that many nurses would prefer to do OBS (observations and note taking) on their own, shown below are the questions we asked different medical professionals, and their responses.

Would you find it helpful if a robot helped with tasks like finding equipment and taking water to patients?
"Yes very helpful to finding equipment but no I like getting my patients water and a blanket because it gives me an opportunity to interact with them."
“When I think about burn out I think of our, physios, mental health workers, pall care workers, pharmacists, & doctors …When I think about automation for any of the above mentioned professions, I think of robots picking medications or checking stores for meds for nurses & pharmacists.”
⁠What if it did obs or did your notes?
"I don't think obs taken by a robot is particularly safe because obs might be ok but the patient looks unwell. Old rule - treat the patient not the obs."
"No I would not like that, I have a method in my chaotic ways. I would find it annoying I think. Kind of like a student nurse, always there and kind of helpful but not autonomous of clinically able to be and you you still have to show them how to do the things etc."
"If I walked into a checkup with a patient with vitals ready that would be a game changer"
⁠Would this influence your choice in which hospital to work in?
We had a mix, with majority people saying yes

What we gathered from our conversations with our target demographic, was that we had a winning product! It was something that was unanimously agreed upon by 10-15 medical workers to be a good idea. What we hadn’t expected was the diversity between people when it comes to observations, and notes. We realized that many nurses prefer to do these kinds of tasks, and it made us think about how we can improve to cater for this. We realized that features should be toggle-able, and if a hospital decides they want to disable a specific feature, or highlight another one, they should have the opportunity to do that. It is important to seek feedback and act upon it when given to you, and if it weren’t for these mixed opinions, we may never have added the customization feature.

Road-map:

Months 1-3:

Gather feedback from medical professionals (done)
Identify values and tasks to automate (done)
Finalize core MVP features (done)
Assemble a small mockup (done)
Get started on building out the MVP (done) Months 4-6:
Develop a line-following robot (work in progress)
Integrate basic voice commands (done)
Delivery Tray Mechanism
Create alpha demo Months 7-9
Pilot Trials in 1-2 sites
Track data such as time saved, as well as nurse satisfaction
Beginning hardware optimization, customizability and upgrades Month 10-12
Pre-Launch
Optimization overall Month 13-15
Set up manufacturing pipeline
Start to negotiate with and create lasting clients
Long Term Goals(4-5 years): Develop better PORTER devices, continue to upgrade Expand to having at least multiple hundred customers Look for expansions into untapped markets

Competitors:

The idea is incredibly niche, and in general only one other place in the world has adapted something this similar. In Japan, they have autonomous hospital robots specifically to select blood samples.

Since the competition is generally secluded to one country, and the fact that they have a lot less features than we do, we are in a generally untapped and uncompetitive market, which we could be the forerunners of.

Risk Mitigation:

Risk: Robot malfunctions in hospital environment Impact: Patient safety issues, workflow disruption Mitigation:

Extensive hardware QA testing
Fail-safe shutdown protocols
Physical bumpers and emergency stop buttons

Risk: Inaccurate navigation or line-following Impact: Inefficiency, blockage of pathways Mitigation:

Combine IR sensors with camera-based SLAM
Test navigation under real-world hospital conditions

Risk: Voice assistant misinterprets commands Impact: Frustrated users, task failures Mitigation:

Use Gemini API with healthcare-focused fine-tuning
Include touchscreen or physical button fallback
Risk: Data loss or software crash -Impact: Downtime, loss of logs or vitals Mitigation:
Regular data backups
Watchdog timers to detect failure
Modular software with component-level restarts

What Inspired Us

Hospitals are complex systems, and nurses often spend a significant portion of their time on non-clinical tasks like fetching equipment, delivering water, or doing clerical work. We came across studies showing that up to 30% of a nurse’s time is spent on tasks that don’t require medical training. That, along with feedback from real nurses across Canada, the U.S., and Australia, motivated us to build something practical, PORTER, a robot assistant designed to take care of these tasks and let nurses focus on patient care.

What We Learned

We learned that automation in healthcare needs to be flexible. While many staff welcomed help with logistics, others preferred to handle patient-facing tasks themselves to maintain human connection. We also saw that designing for real environments like hospitals means balancing convenience, reliability, and safety. The biggest takeaway was to always build around user needs, not just technical capability.

How We Built It

PORTER is a line-following robot that uses:

IR sensors for track following
Ultrasonic sensors and cameras for basic obstacle detection
A microphone and speaker to support voice commands
Gemini API integration for natural language understanding
A Python backend for handling commands and coordinating navigation
A custom control dashboard to manage routes, settings, and permissions

Our goal was to make it simple, modular, and easy to deploy in hospitals or clinics.

Challenges We Faced

The biggest hurdles were on the hardware side. We initially chose the ESP32 for its Wi-Fi and Bluetooth support, but ran into limitations around processing power. It was hard to keep the robot lightweight yet powerful, especially when trying to integrate camera feeds, real-time sensors, and voice recognition.

Another major issue was that many available libraries and sensor headers were written in C, while our control logic and bot code were written in Python. This mismatch meant we had to spend extra time on integration, debugging low-level behavior, and rewriting parts of the system to ensure compatibility.

Despite these challenges, we created a working prototype that reflects our vision: a helpful, voice-enabled assistant that can reduce nurse burnout and increase efficiency in healthcare environments.

Sources:

Nature Portfolio. (2024, June 14). Robotic hospital assistants that can independently transport medicine and documents. Nature. https://www.nature.com/articles/d42473-024-00343-5

TrueSpot. (n.d.). Unlocking the real ROI of RTLS asset tracking in healthcare. TrueSpot Health Blog. http://truespot.com/healthblog/unlocking-the-real-roi-of-rtls-asset-tracking-in-healthcare

Cornell, P., Gervis, M. T., Yates, L., & Vardaman, J. M. (2016). Nurses' time allocation and multitasking of nursing activities: A time motion study. Nursing Research and Practice, 2016, Article ID 5729018. https://scispace.com/papers/nurses-time-allocation-and-multitasking-of-nursing-4z6pm9bjh6

Health Canada. (2022, June 6). Reduced administrative burden. Government of Canada. https://www.canada.ca/en/health-canada/services/health-care-system/health-human-resources/nursing-retention-toolkit-improving-working-lives-nurses/reduced-administrative-burden.html

Hendrich, A., Chow, M. P., Skierczynski, B. A., & Lu, Z. (2008). A 36-hospital time and motion study: How do medical-surgical nurses spend their time? The Permanente Journal, 12(3), 25–34. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3037121/

24x7 Magazine. (2018, May 17). Nurses searching for equipment led to $1.4B in lost productivity. https://24x7mag.com/management/equipment-management/nurses-searching-for-equipment-led-to-14b-in-lost-productivity/

Canadian HR Reporter. (2006, June 5). Let nurses do their jobs: Report. https://www.hrreporter.com/news/hr-news/let-nurses-do-their-jobs-report/311744

Potter, P., Boxerman, S., Wolf, L., Marshall, J., Grayson, D., Sledge, J., & Evanoff, B. (2004). Mapping the nursing process: A new approach for understanding the work of nursing. Journal of Nursing Administration, 34(2), 101–109. https://pubmed.ncbi.nlm.nih.gov/23312466/

Canadian Federation of Nurses Unions. (2022). Sustaining Nursing in Canada: A Call to Action. https://nursesunions.ca/wp-content/uploads/2022/11/CHWN-CFNU-Report_-Sustaining-Nursing-in-Canada2022_web.pdf

McHugh, M. D., Berez, J., & Small, D. S. (2013). Measuring nursing care time and tasks: A new tool for research and practice. Journal of Nursing Care Quality, 28(2), 135–144. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6171224/

University of Alberta. (2020, January 29). Study finds most nursing home aides report missing or rushing essential care tasks. Folio. https://www.ualberta.ca/en/folio/2020/01/study-finds-most-nursing-home-aides-report-missing-or-rushing-essential-care-tasks.html

Indeed. (n.d.). Registered nurse salary in United States. https://www.indeed.com/career/registered-nurse/salaries

Indeed. (n.d.). Registered nurse salary in Canada. https://ca.indeed.com/career/registered-nurse/salaries

Herzing College. (n.d.). LPN salary guide. https://www.herzing.edu/salary/lpn

Indeed. (n.d.). Licensed practical nurse salary in Canada. https://ca.indeed.com/career/licensed-practical-nurse/salaries

Indeed. (n.d.). Nursing assistant salary in United States. https://www.indeed.com/career/nursing-assistant/salaries

Timmerman, M. (2021). Nurses’ time allocation and multitasking of nursing activities: A time motion study. Academia.edu. https://www.academia.edu/85676993/Nurses_Time_Allocation_and_Multitasking_of_Nursing_Activities_A_Time_Motion_Study