Experimental setup for 1 dimensional constrained frictionless motion Physics textbooks are full of examples considering one dimensional constrained motion (in the gravitational field) with neglectible friction.
However I have never seen setups with curved paths where friction is really neglectible. There are non curved air tracks out there with neglectible friction and experiments with curved paths like this roller coaster model but with non neglectible friction. To make the question more concrete, consider the latter example: 

I know this experiment and if you try to show conservation of energy (for example by measuring the top height and speed at the bottom) you have losses of more than 10 percent (depending on the track shape even much more).
So, what is the state of art to make such experiments with as low friction as possible with a curved track (like in the lower track of the image above)?
 A: I'm not an "expert" on reducing friction losses, but I think you need to consider viscous friction (due to contact of the vehicle with air, i.e., air drag) and dry friction (due to contact between the vehicle and the track, and the friction of the car axles).
As far as air drag is concerned, you could probably benefit from the designs of aircraft and motor vehicles. Although the scale is different, designers probably now what those factors are when they use wind tunnels to design the contours of the vehicle.
As far as dry friction is concerned, you need to look at the wheels of the vehicle and the type of surface of the track. Since gravity is the force propelling the vehicle (as opposed to wheel torque and static friction) I would think you would look for track surfaces that have the lowest coefficients of friction. 
Then there are the materials and design of the wheel surfaces and the "axles". I remember working with my son on cub scout car racing and the benefit of using graphite on the wheel axles to reduce axle friction.
These are only some observations. They are not obviously mathematically rigorous, but perhaps they can help guide the analysis.
Regarding the use of magnetic levitation, as has been suggested, correct me if I'm wrong, but aren't you looking for ways to minimize lap time without the presence of  forces other than that of gravity and friction?
I think air drag is a minor error in such experiments
You're probably right. You might consider a simple experiment to prove it. Configure a vehicle to try and deliberately introduce air drag without influencing other factors. Perhaps install a lightweight "sail" on a vehicle and see if it measurably slows it down.  
One last major consideration. Are there any constraints on the design of the vehicles on the track. I mean, if there weren't any you could just use stainless steel balls as @user9976437 suggested. If this is a competition, then there must be some rules. And that goes for magnetic levitation as well. 
So I guess the first question any of us should have asked you before answering your question is what are the rules of the competition, if any. 
Hope this helps.
A: Basically, you need to reduce the influence of friction.


*

*from air. (It did matter if you rail was as large  as shown in your picture, especially when it involved $v^2$ term. We did the lab in sophomore year, the $v^2$ term and $v^3$ term will start to show up with that kind of volocity) {answer: round ball}

*from rail. (Avoid materials that's sticky.){answer: use soft oil to cover the rail.}
Another factor to take into account was the density, consider ping pong ball.
So A. a cheap way to do it was with a good size stainless steel ball.
However, there's yet another option if you got the  budget and time B. use magnetic rail and superconductor. that way the only thing you need to worry about was air resistance.{and what's best, magnetic does no work}
