The reality is that older population worldwide is expected to grow to 2 billion by 2050, according to United Nations report. In US alone, there are 76.4 million Baby Boomers (those born between 1946- 1964) that are ageing as we speak. Other than the progressing health needs, this group is facing a serious issue of loneliness. According to a recent study by researchers at the University of California, San Francisco, 18% of seniors in US live alone, while 43% report feeling lonely on a regular basis. Lonely seniors are 45% more likely to die faster. With that in mind we came up with LISA, short for Logical, Intelligent, Special Assistant. It is a personal companion and caregiver for the elderly
What does it do?
LISA would allow its client to live an independent and healthy life. LISA can track its client by using an opencv program and a webcam on LISA, that can track a certain colour. The opencv program can tell where the object is on the screen, differentiating whether the object is on the left side, right side and centre. By doing this, LISA can tell whether it has to turn to make sure the object is still in its line of sight. It would send a signal to the Arduino that controls the main servo, telling it to turn in the correct direction. Furthermore, a secondary Arduino controls the motors of the chassis. An RF key fob sends signals to the secondary Arduino to move the tracks on the chassis. This give the robot a wider range of movement, allowing it to not only rotate on the spot, but also move around a room. Another key feature of LISA is its ability to remind you of different events. For example, LISA could remind you when to take your medication. It would use our medication dispenser to dispense the medication. It uses a ZX distance and gesture sensor to detect hand movement toward the pill dispenser. When this movement is detected, LISA will dispense the right amount of medication. Moreover, LISA would do regular checkups of the client's heart rate, and based on that, would suggest recommendations. For instance, if the heart rate is too low, LISA might recommend for you to take a walk around the park. Or if your heart rate is too high, LISA might recommend for you to take a break and rest.
Because we wanted LISA to have a real personality instead of being just a screen, LISA would communicate with its clients by talking. When LISA has to tell you something, it would talk to you instead of just displaying a message. On top of that, LISA would occasionally tell jokes to make sure that the client doesn't get bored.
How we built it
The main chassis for LISA is from a starter robot kit by Makeblock. It has two motors and four wheels, moving two tracks like a tank. We also used a cylinder made out of cardboard to elevate the base of the laptop. An acrylic platform was used to create a larger surface area for the laptop. Plastic trays were mounted around the robot to store most of the electronics (Arduinos, battery packs, sensors, etc.). A lid of a laundry detergent was used as a container for the medication.
Challenges we ran into
A large challenge we ran into during our robot build was with the system of the interfacing of the object tracking and movement. The robot would not be tracking an object when it was making the movement of 'searching' for it. So, instead of having the robot sweep back and forth in one motion while looking for an object, it would make smaller intervals in between. It would then stop in between each interval. So, when an object is not in view, the robot would rotate a bit, stop, search, rotate, and repeat. This way, we are increasing the refresh rate of the searching.
Accomplishments that we're proud of
Object Tracking We're really proud of the object tracking because we were able to get it functioning properly in a very short period of time. Furthermore, the movement of the track was very accurate, so that the camera is always looking in the direction of the object.
Medication Dispenser We're proud of the medication dispenser because we used a sensor that we've never used before, the ZX distance and gesture sensor. It took a very long time to get the gesture working with the servo to dispense the medication. We found out that the sensor has to be facing up towards light for the sensor to work 100 percent of the time.
Speech Speech was great and really fun to do. We were lucky that we were able to find a website with a text to speech feature with a natural sounding voice. The outcome of the application of the speech is awesome because it gives the robot a personality and a voice. It allows the robot
What we learned
We learned how to track objects, text to speech, speech interfacing with object tracking, RF communication, how H bridges work and how to use the gesture sensor.
What's next for Logical Intelligent Special Assistant [LISA]
Some next steps for LISA would be to create an app so that the relatives or kids of the client can check up on the robot and get a look at what the client has been up to. Another next step would also be facial recognition instead of colour recognition to identify the client. Finally, we would also want LISA to be compressed into a smaller form factor so that its not obtrusive. It can also use beacon technology to follow you. It can also be able to use speech recognition to communicate with the robot. The client's doctor would also be able to access LISA to access the client's information. Other sensors would be applied in the future, such as blood pressure and oxygen sensor.