[Use Case TF] Scheduling medical treatment (from Berners-Lee et al., 2001)

[scranefield]

Title: Scheduling medical treatment (from Berners-Lee et al, 2001)

Submitter(s):

Stephen Cranefield

Description:

This use case is from the article "The Semantic Web" by Tim Berners-Lee, James Hendler and Ora Lassila, Scientific American, 2001 (https://lassila.org/publications/2001/SciAm.pdf).

A series of medical appointments need to made. The providers and appointment times must be chosen based on multiple constraints:

• the providers must be approved by the patient's insurance company and be well rated on trusted rating services,
• the treatment locations should be within a certain distance of the patient's home,
• the appointments must match the schedules of the patient and family members who will provide transport (with rescheduling of existing activities considered if necessary), and
• the patient and family should be consulted before finalising bookings to allow other preferences to be taken into account (e.g. avoiding rush hour traffic).

Expected Participating Entities:

Web sites, Web services, software agents.

Illustrative scenario(s):

Lucy has just taken her mother to see a doctor, who says that a specialist appointment is needed and then a series of physical therapy sessions. Lucy asks her assistant agent to set up the appointments. Lucy also phones her brother, Pete, who agrees to share the chauffeuring duties. The scenario proceeds as follows (this is taken verbatim from the Scientific American article):

At the doctor’s office, Lucy instructed her Semantic Web agent through her handheld Web browser. The agent promptly retrieved information about Mom’s prescribed treatment from the doctor’s agent, looked up several lists of providers, and checked for the ones in-plan for Mom’s insurance within a 20-mile radius of her home and with a rating of excellent or very good on trusted rating services. It then began trying to find a match between available appointment times (supplied by the agents of individual providers through their Web sites) and Pete’s and Lucy’s busy schedules. ...

In a few minutes the agent presented them with a plan. Pete didn’t like it— University Hospital was all the way across town from Mom’s place, and he’d be driving back in the middle of rush hour. He set his own agent to redo the search with stricter preferences about location and time. Lucy’s agent, having complete trust in Pete’s agent in the context of the present task, automatically assisted by supplying access certificates and shortcuts to the data it had already sorted through.

Almost instantly the new plan was presented: a much closer clinic and earlier times — but there were two warning notes. First, Pete would have to reschedule a couple of his less important appointments. He checked what they were - not a problem. The other was something about the insurance company’s list failing to include this provider under physical therapists: “Service type and insurance plan status securely verified by other means,” the agent reassured him. “(Details?)”

Lucy registered her assent at about the same moment Pete was muttering, “Spare me the details,” and it was all set. (Of course, Pete couldn’t resist the details and later that night had his agent explain how it had found that provider even though it wasn’t on the proper list.)

The article proposes that reasoning with ontologies will be an important part of the solution, although this is not illustrated in the scenario description above.

The importance of Web Agents for the use case

This scenario involves a task that requires interactions with multiple web-based resources and that involves coordinating activities of multiple people. This is exactly the type of task that software agents were originally proposed to help with. Automation of tasks like this by intelligent agents with high level web interaction skills and planning capabilities promises to provide huge efficiency gains to people in both their work and daily life.

Existing solutions

More than 20 years since this scenario was published, there is still no complete solution that I know of.

Other information (optional)

N/A

[Antoine-Zimmermann]

@scranefield Belated thanks for the use case. This is definitely a fit for a Web Agents use case. What I'm wondering is to what extent this is doable today. Let me list things that are required for this to work:

• The assistant agent must have knowledge about calendar information. There are standards for calendar information, so it would be doable to identify what slots are available. However, there are tricky situations: when someone has a thing in the calendar but they have the option to cancel, or they are just recording the exitence of an event that they do not necessarily plan to attend. There are also available slots during the night, which should not be used, probably; or at lunch; at dinner. Of course, the slot must match the specialist availabilities, but if we imagine a teleconsultation with a doctor in a different time zone, then the doctor's availability may match the patient's night. Probably the agent would have to make simplifying assumptions.
• The agent must know about the type of service that it needs to retrieve. If the specialist that must be consulted is a dentist, the agent must not select a pediatrist. But an orthodontist would fit as well. Today, schema.org has a good coverage of medical businesses and their types, so we can assume it is doable if such services are correctly annotated on the Web.
• The agent must know about ratings. This is also well covered by schema.org, but ratings do not necessarily indicates trustworthiness.
• The agent must know about geolocations. This is well covered by today's technologies.
• The agent must negotiate with the providers' agents, or at least with a service that allows the agent to access the providers' calendar, but only the availabilities, not the details. There may be a need to enforce privacy policies.
• The assistant must be able to do some things autonomously (such as contacting the right data sources or stakeholders) but must also report to their owner and wait for approval sometimes. The agent must have a plan of actions to perform so as to gather the right information at the right time.
• To be more flexible and be useful in several situations of different nature, the agent should be able to adapt to a process description rather than having all tasks and plans hardcoded.
• The assistant must be able to access other people's schedule that are relative to its owner (family members), again with privacy awareness.

While some of these things are already standard (e.g. calendar descripiton), there is a lot of infrastructure to put in place for this to work. Not only the right knowledge must be available in the right form at the right place, but also all stakeholders must act according to a common plan in everybody's interest. If some knowledge is missing or incorrect, or one of the actor is misbehaving, then the system should be able to gracefully fail.

For these reasons, I'm wondering if there could be a simplified version of the use case that would be more or less implemented in order to be demonstrable.