Enrico Gastaldelli
đŸ‡§đŸ‡·đŸ‡ș🇾

Case 02 · UX Research

Next

User research and personas for an IoT solution that monitors medication temperature in hospitals.

Client
Next, an IoT company (name kept confidential)
Year
2019
Role
User research and persona designer
Services
UX Research
Interviews and surveys
Empathy map
Personas
Interaction design
UI design
Team
Enrico Gastaldelli
Alessandro Narimatsu de Faria
Multidisciplinary team of 12 people
Links
Behance
NEXT logo in white on a black background.
01

About Next

Next was created to reduce the strain of manual work and processes, redirecting that time to activities that really matter and giving its customers peace of mind by connecting them in a simple, intuitive and smart way.

Today Next's portfolio includes a range of IoT and technology applications based on tracking and managing objects with RFID and Smart Tags. These solutions let customers monitor any object with a small tag equipped with sensors that capture many kinds of information, such as real-time temperature and humidity of equipment and rooms, without conventional thermo-hygrometers or paper logs.

02

Context

The digital transformation of healthcare started years ago, but the main advances have been in diagnostic and surgical equipment.

Countless processes and procedures still rely entirely on manual work, prone to human error and very time-consuming.

  • R$ 1 billionin medication wasted every year in Brazil
  • 6,7%of Brazilian hospitals report significant losses of supplies due to temperature deviations in storage
  • US$ 35 billionin medication losses in the cold chain in 2015, according to an estimate by CargoSense

Temperature is the most important factor in the degradation of pharmaceutical products: as temperature rises, chemical degradation increases.

03

Goals

  • Describe the product persona. Research users and the market to define the best platform for consuming data and using the system (mobile or desktop), based on engagement and the customer's context.
  • For the mobile application, gather data showing whether users prefer a native app (iOS, Android etc.) or whether accessing the platform on the mobile web is enough, given the system's range of features.
  • Conclude which platform and which UX are best for the product in the healthcare market, based on data collected from the market itself, from users and from customers.
04

Process

From research to application, the process was split into 4 stages.

  1. 01

    Exploratory research and hypotheses

    Understand which goals Next could reach with the current product, get to know potential users and the routine of the professional in charge of temperature control, and write an interview script based on the scenario found in desk research.

  2. 02

    User interviews

    In person, by phone and through an online survey, refining the questionnaire and visiting the workplace.

  3. 03

    Defining the persona and the platform

  4. 04

    Application

    What kind of technology and service Next could offer with the product in its expansion phase.

05

Hypotheses

Desk research surfaced a few possible problems.


healthcare systems that don't invest in technology usually show several inefficiencies


The solutions on the market seem to follow a common-sense pattern of what needs to be done: control temperature and humidity, prevent bacteria from spreading, “have great air conditioning” and automate processes. Overall, different technology companies are offering the same answer.

06

Interviews

The first reaction was positive: we introduced ourselves as students and explained how the program worked. But once the topic came up, the mood changed completely and the interviewees became very wary and suspicious.

We're told not to talk to companies.

Coordinator at one of the hospitals, who refused to let the pharmacist take part in the research

In-person interviews were held at 2 hospitals, and the equipment situation was the same in both.

Both interviews were with women, pharmacists with postgraduate degrees. The main pain points they raised were the lack of investment in equipment and the absence of staff at the unit on weekends, which makes power outages a big problem.

  • 01

    Fernanda, 31

    Pharmacist, single.

    Pain points: power outages, equipment quality, audible alarms that don't say what caused them, and the unit being closed for long periods.

    Goals: a manual with guidance on what to do in each situation (a flowchart), reports as needed and a thermometer with exact readings.

  • 02

    Juliana, 40

    Pharmacist, divorced.

    Pain points: lack of investment in technology, makeshift equipment, a hectic routine and no temperature control on weekends.

    Goals: the right fridge, automated work and automatically generated reports.

If there's a power outage on the weekend, we only find out on Monday.

The interviews confirmed that hospitals don't invest in technology (source) and that they are selective about where they do: most of the time, the money goes where the hospital could be fined by ANVISA, Brazil's health regulator.

07

Survey research

We collected 18 responses on Google Forms, 2 of them by phone. A second survey was needed to follow up on answers from the first form.

Both surveys had 2 sections. The first covered personal details (name, age, education, job title, marital status
). The second was aimed at building the empathy map and the persona's story (hobbies, habits, family, friends, personal goals).

  • 13responses to the first survey
  • 5responses to the second survey

The empathy map brought together what professionals think, hear, see and do in temperature control.

  • 01

    Thinks and feels

    I feel powerless, because I've asked for the right equipment many times.

    I feel control should be stricter.

    Frustrated: management isn't interested.

    Changes could make professionals' work easier.

  • 02

    Hears

    Lack of budget.

    Trouble identifying what set off the alarms.

    The whole team agrees on what the ideal would be.

    There's a lot of pressure around the checks.

  • 03

    Sees

    Makeshift thermometers.

    A search for technology-based solutions.

    Lots of copy and paste, especially in internal logs.

    Market guidance isn't explicit about storage.

    We always have to keep an eye on the thermometer.

  • 04

    Says and does

    There isn't much to do except report the situation.

    I try to make the team's work easier by encouraging immediate communication.

    I do it however I can.

    I pass the information on when there's guidance.

Overall, users' pain points are mostly tied to power outages.

Power outages, lack of maintenance.

Equipment quality, power outages and maintenance.

Power outages, temperatures that aren't always accurate.

  • 61%point to power outages as the main problem

Other open answers about pain points:

  • Power outages, which are frequent in many places, and the very poor quality of the thermo-hygrometer, which can fluctuate or even show the wrong temperature.
  • Audible alarms that often don't let you identify the cause. In a unit that doesn't run 24 hours, the team only learns about the outage at the start of the shift: how high the temperature went, what time the power went out, whether the medication's quality is assured. Often only the security guard is at the unit.
  • All of these, plus transport logistics.
  • Lack of backup systems for routines or research that depend on power, and equipment stopping suddenly mid-process.

One answer stands out among the pain points.

I worried about the patient; many times I gave the medication afraid of how they would react.

Besides the equipment problems, we also see the stress professionals face in a setting that favors neither innovation nor a healthy work culture.

08

Personas

The personas were split into two categories: the management profile and the operational profile.

The management profile covers professionals in senior roles, who make decisions and influence the purchase and rollout of new technology. They aren't in direct contact with temperature control, but they answer for it if something goes wrong.

  • 01

    Daniel, 40

    Pharmacist, married, 2 children. Quality assurance manager.

    Routine: 15 years in the pharmaceutical industry, 5 as a manager. Spends 2.5 hours a day in the car commuting. Goes to parks and malls with his family and isn't a heavy social media user.

    Work: tries to automate the department's processes to reduce human interference and errors in the data. Keeps up to date through technical websites, events, conferences and union meetings.

  • 02

    Marcela, 37

    Biomedical scientist, single. Clinical laboratory manager.

    Routine: 8 years in healthcare, 2 as manager of a hospital's clinical laboratory. Spends 1 hour a day in the car. Runs street races, travels on holidays and posts on Instagram, Facebook and LinkedIn.

    Work: pursues continuous process improvement, fewer manual tasks and online data management. Keeps up to date through journals, conferences and scientific papers.

The operational profile is made up of those on the “front line” of temperature logging, control and storage. They have little decision-making power and, in team meetings, often complain about the quality and importance of the equipment. When asked about investment in their workplace, these professionals are clearly frustrated.

  • 01

    Rafaela, 42

    Pharmacist, single.

    Routine: 12 years as a pharmacist, lives in the city where she works, has a postgraduate degree and balances friends and family, with a focus on work and study.

    Work: runs into budget limits and a lack of innovation culture when trying to bring in technology. The biggest problem is fridge quality and power outages. She raises these issues in meetings with supervisors and managers.

  • 02

    SĂ©rgio, 30

    Nursing technician, single.

    Routine: 5 years as a nursing technician, in public and private healthcare. Lives in the next town, goes out with friends after work and is a heavy user of Facebook and Instagram.

    Work: makes do with the equipment provided, struggles to keep up the temperature control routine and has no autonomy to bring in technology. He knows that slip-ups affect medication quality.

At first, the research pointed only to the pharmacist profile (in general). As the work went on, though, we found that the cold chain has other professionals who handle temperature-sensitive medication and could also benefit from Next and its product.

09

Application

After the second industrial revolution, with mass production, and the digital revolution, which automated processes, Industry 4.0 emerges as the natural path to make the sector more competitive through digital technologies that cover automation and data exchange, with concepts such as Machine Learning, IoT and Cloud Computing.

Next's current system shows the same information as other players on the market, and it is also reactive, with alerts only after a deviation happens.

In this innovation project, we suggested moving to an active solution, in line with Industry 4.0, where one more piece of information is added to what users monitor: the probability that the temperature will go beyond the set limits. That way the solution stops being reactive and makes preventive action possible.

Development would happen in Next's system, adding a Machine Learning tool. The hardware would stay the same.

  • 01

    Alerts

    Acting before the incident happens, reducing the risk of product non-compliance.

  • 02

    Risks

    Seeing the probability of the temperature leaving the quality range.

  • 03

    Intelligence

    Machine Learning applied to big data, correlating sensors, historical data, external factors and statistics.

The system would analyze the big data of sensor readings and timestamps, correlating data from devices in the same room, including temperature, humidity and city power outage variables, comparisons with sensors in other locations, statistics and an algorithm. This makes active decision-making possible, reducing loss rates, ensuring reliable quality and effectiveness of the products given to patients, and protecting the hospital's or clinic's reputation.

10

Interface

Based on the discovery, I designed 3 dashboard screens for the active prevention proposal: one for managers, one for whoever operates the equipment and one for the incident that used to be discovered only on Monday.

Overview dashboard: indicators at the top, a predicted risk table with likelihood and probable cause, a list of departments and a banner about the weekend shift.
01 Overview (management: Daniel and Marcela)
  • Indicators at the top: how many devices are in range, how many are at risk in the next 6 hours, open alarms and compliance over the last 30 days.
  • Active prevention: a “Predicted risk” table lists the devices still in range that are likely to leave it, with the likelihood and probable cause. This is the core of the case: warn before the deviation, not after.
  • Departments and on-call: the status of each department and a banner about the weekend, with the schedule of who gets notified and through which channel.
Device screen: the day's chart with the 2 to 8 °C range, the reading, a dashed forecast and door openings; next to it, a step-by-step of what to do and who is being notified.
02 Device (operational: Rafaela)
  • The day's chart: the 2 to 8 °C range, the actual reading, the dashed forecast with its margin of uncertainty and the moments the door was opened.
  • Cause explained: a short text says why the risk went up. It answers the “the alarm doesn't say what caused it” pain point.
  • “What to do now”: a step-by-step guide, which was Fernanda's goal (a manual for what to do in each situation), and the list of who is being notified and who steps in if nobody acts.
Incident screen: timeline of the power outage in the early hours of Saturday, what happens to the affected batches with a double check, and an audit report.
03 Incident (the weekend power outage that was only discovered on Monday)
  • Overnight timeline: power outage, alerts, escalation to the next level, temperature passing 8 °C, power coming back and the return to range.
  • Batch outcome: discard, check with the manufacturer or release, with a double check.
  • Audit report: generated in the format required by Anvisa.
11

Team

The case study had a multidisciplinary team of 12 people, including 2 designers: Enrico Gastaldelli and Alessandro Narimatsu de Faria.

12

Sources