First, I am assuming that any starship will take the slow-boat approach to Alpha Centauri, or wherever we decide to send it. Between now and then, solar-system based technology such as precision telescopes and radio astronomy, will have helped us know more about nearby star systems so that we can choose the one most likely to support some sort of human colonization effort, assuming that is the ultimate goal. It is probable too that before building a manned starship, we will have sent robotic ones to nearby stars by some method; enough to have explored them somewhat, so that we are not sending people in blind. That would presumably take several decades, but would also serve as preliminary work for the Starship development itself.
The slow-boat approach assumes no "warp drive", magic gravity-changing techniques, or other speculative ways to achieve non-reactive thrust and thus, sustained high acceleration. The slow boat is just a continuous push at low acceleration, over a long time (decades), something we can mostly understand today, with no new physics required. See more below on Starship drive possibilities.
Biotechnology will be one key component for sending people and other life forms to the stars with minimal weight penalty, little life support requirements, and no worries about boredom en route. We are already developing the means to go from DNA to living lifeforms in the lab. With further development, it should eventually be possible to reliably reconstitute many different animals and even humans from a DNA library. Plants can be grown from seeds, and in some cases, animals can be transported as frozen embryos. We already have that capability for simple lifeforms, and we can store human ova and sperm, or embryos for years. Future technology may approach artificial-womb status, at least for animals, allowing various species to be regenerated at the far end of the voyage.
The ultimate would be an automated laboratory that can generate any DNA from a data file and then use biochemical and cellular mechanics to grow each lifeform from its particular DNA strand. Using such an approach, a Starship could carry a significant population of numerous species - even an entire biome - to another star system, with minimal mass/volume and life support costs. The laboratory I have in mind does not yet exist, of course, it would have to be highly automated, and reliable enough to work properly after decades in space - a major hurdle for something so complex. Some sort of advanced AI and general purpose robotics would be needed to do most of the work and render the lab operational after the long flight.
To maximize the chance of a successful trip, a Starship will most likely need some live, trained humans at some point during the voyage, to handle unexpected emergencies, or at least to kick start and supervise activities upon arrival. Therefore it may be good to send a few (perhaps six?) actual human adults along in some sort of yet-to-be-developed cryo-sleep. I envisage a technique for greatly slowing human metabolism, supported by blood additives and a variety of chemical drips, cooling down the human body to perhaps 10°C. The human would then be in a coma, monitored by the robotic AI, and could remain in that state for years with perhaps little aging.
There has been some work in this area, but it obviously needs a lot more research to become practical, especially for reawakening the person at the end of the trip, or if needed earlier. I assume the robotic AI would be in charge of the Starship through most of its voyage. Awakening one or more humans only when needed at the end, or perhaps in an emergency. The human crew would likely be all female so that they can have children (via in-vitro fertilization) in the new world. Biology is not destiny, but it helps in this situation.
To minimize travel time, the starship would likely have to get up to somewhat relativistic speeds midway through the voyage. Thus, the people and stored DNA inside would be subjected to increasing, and increasingly directional, ionizing radiation in interstellar space: high energy photons, electrons, neutrons, protons and an occasional atomic nucleus. To shield the biological "cargo" against radiation damage over decades, heavy water could be used to surround the sensitive humans and other biological specimens. Of course, the same heavy water could provide the fuel for a fusion drive for the Starship.
Ultimately, of course, the primary limitation on any Starship design is the drive; how to get decades of significant acceleration out of an engine, using less than 100% of the starship mass for fuel. At present, if I may be so bold, the only significant hope here is fusion power. Granting that any practical fusion generating station is many decades off (see this for an explanation), the concept of directed fusion exhaust pushing a starship with some reasonable efficiency, is not impossible, at least in principle. No, we cannot do it today, but maybe in a hundred years we will be able to harness the materials, physics and other technologies needed to do so reliably and sustainably.
The idea is to take tons of heavy water (D2O), break it down, and use the deuterium atoms to fuse into helium nuclei, yielding enough energy to blast the helium, loose electrons, and perhaps the left over oxygen out the back of the ship, at high speed, more or less in one direction. Fusion energy gives the biggest push per unit mass of fuel that we are likely to see for a very long time. The starship would be built in orbit and then accelerated at some low but steady rate until halfway through the voyage. It would then turn around and decelerate the rest of the way so that it could arrive at a low enough velocity to successfully enter the target system and orbit a planet there.
What realistic acceleration a fusion drive would yield and what ratio of fuel mass to ship and cargo mass it would need to travel a few light years, is for others to calculate. They have done so under varying assumptions, and although not a slam dunk obvious success, there are some hopes of eventual good results. Aside from the technical hurdles of achieving continuous fusion, keeping the torch burning successfully for decades would also require major engineering design work. Nothing in Starship design is easy!
I have mentioned robotic AI above, and that is a more promising area of development. There have already been major advances in robotics and artificial intelligence, and these will doubtless continue and become more advanced over the next century. While I doubt we will ever make a human-like AI system, the software will surely advance to the point that most controls and processes can be fully automated. There would have to be numerous repair mechanisms, contingencies and reset protocols built in, along with considerable redundancy, but we humans already do a lot of that for other purposes. Fortunately, AI software does not take up much space or power, and robots needs no life support systems.
There would also be major technology requirements at the end of the trip. What needs to be done when the Starship has arrived at its destination and entered orbit around a selected planet? The human crew (or cargo) would have to be awakened from their deep sleep, and allowed to recover. The AI could continue operating the Starship, and could perform much of the lab work and orbital tasks before anyone or anything actually goes down to the planet. The fusion drive could be shut down or turned way down to provide ongoing electrical power.
Previous, unmanned missions would most likely have occurred to test the Starship technology, and to deliver most of the hardware needed by the crew at their destination. Landing shuttles and their fuel, supplies for the crew, materials for on-orbit construction, research and communications equipment, extra habitat modules, and so on could be delivered in advance of the manned Starship. These earlier missions would also be scouting out the system for potential habitations such as asteroids, moons, or "goldilocks" planets, and collecting long term data to minimize surprises after the ship's arrival. Thus the crew would have a head start via robotics at the far end, and their ship would only finalize its approach once everything was reported to be in place and ready. Moreover, they would likely stay on orbit for months in preparation for descent or major construction projects.
After arrival, detailed plans would be made to begin colonization. The crew, aided by the robots, would prepare needed equipment, likely by cannibalizing the Starship materials, which would have been designed partly for that purpose. The (female) crew would become pregnant (IVF), deliver babies occasionally, with some eventually being males. If artificial wombs are then possible, they could be put into service instead. The crew and support robots would grow or synthesize food, perhaps initially using algae and yeast cultures on orbit. Most supplies can be recycled, and materials could be reused or remodelled into needed equipment or tools, probably using some form of 3D printing. Similar to the plans afoot at NASA to set up a colony on Mars, materials could be used for a larger habitat. With a suitable habitat, they could grow plants and start to generate live animals.
Eventually, the colony would want to descend to the surface. For that, they would need to have studied the planet in detail from orbit, and made plans and contingencies for every scenario. Probably only a small team, complete with robotics would be sent down initially to get things going, build the infrastructure and test out the processes for living there. It is unlikely the planet would have abundant life and an oxygen atmosphere, so processes and equipment for life support would have to be made or taken down from orbit. The planet-bound people would work on developing soil, water supply, power, waste management, and all the myriad other things needed to support a small colony, using mostly materials extracted from the planet. None of this will be easy, and people will probably die along the way, but none of it is impossible in principle.
Once the initial colonists have shown they can live on the surface, grow food, recycle effectively, and generally make a go of it - possible over a year or more - the remaining people on orbit, and the DNA database and labs could be brought down as needed, and the colony would then grow slowly and carefully. Whether some crew remain on their now demoted Starship in orbit, or they leave that up to the robotic AI, can be decided at that time. However, the idea is that there should be no need for people to return from the surface back to the Starship, thus simplifying the colony transportation requirements (no huge rockets needed).
Aside from the fusion drive and the futuristic human biotech, the other major hurdle will probably be reliability. How can we ensure the AI, robotics, and bio-lab will continue to function, or can be self-maintaining over decades in the unforgiving environment of interstellar space? How could we maintain complex systems without human attention over the same period? Given the track record of computers, complex factories, and mechanical hardware, even here on Earth, the prospects are daunting. This calls for careful engineering and probably redundant design at each level, up to and including multiple Starships to same destination. Having a human crew available during the trip might improve the chances of success, but would entail other requirements and raise other problems, such as life support and human sanity over decades in space.
Serious science fiction writers have grappled with these issues and suggested future technologies along these lines for a long time. As technology advances and our understanding of physical processes and the requirements for space travel improve, the fiction becomes more detailed, and in some cases, more believable. At some point in the distant future, perhaps a century from now, the science fiction will turn into science reality, and serious engineering can begin to design a real Starship, perhaps using some of the above concepts. By then, mankind will have learned to live on Mars, and possibly other planets or asteroids, and nearby star systems will have been studied in great detail. By that point, sending off a Starship to seed the galaxy will not seem so far fetched, but will simply be the next giant leap for mankind.
Addendum: Starship Drive Options
Some further notes about potential starship drive options. The energy needed to accelerate a mass M to a velocity v is E = 0.5Mv^2. Ignoring relativistic effects for now, at v = c the speed of light, that would become E = 0.5Mc^2. Since the starship has to slow down at the other end, a similar amount of energy would be needed, yielding E = 2x0.5Mc^2, or E = Mc^2, which of course is Einstein's formula for the total energy in a given mass. Relativistic effects make this worse since the mass increases as the velocity approaches c. Therefore, to get a starship up to anything close to c and back down to zero would require converting the entire starship mass M into kinetic energy E at 100% efficiency. That of course is impossible.
So without some sort of "warp drive", a starship using only the fuel it carries cannot achieve truly relativistic speeds. If we limit ourselves to fusion, using heavy water as fuel (as mentioned above), then the maximum speed would be considerably lower. The fusing of deuterium to helium nuclei converts up to 5% of the mass into energy. That energy, with some losses, needs to be turned into kinetic energy of the exhaust that pushes the ship forward. Doing so heats up the oxygen atoms (four times the weight of the deuterium) as part of the exhaust, without adding more energy. For a given mass of heavy water then, the most energy you can get by fusion is probably less than 1% of the total mass equivalent.
Assuming that, say 90% of the initial mass of the ship is heavy water (a rather unlikely extreme), the mass converted to velocity to get up to speed must be less than 0.9x0.01x0.5xM = 0.0045 M. (The other 0.5 is needed to slow down.) Applying this in the E = 0.5xMv^2 equation yields a maximum velocity v around 0.067 c, or less than 7% the speed of light - not really relativistic. The average speed (from zero to this maximum) is then perhaps 3.3% of c. At that rate, it would take the ship 130 years to get to Alpha Centauri, the nearest star! Definitely a slow-boat prospect! The ship could probably do a bit better since its mass would be greatly reduced at the turn-over point due to the loss of the fuel used to get there, thereby requiring less energy to slow it down. Nevertheless, the ship would not get much faster than the 3.3% of c and the total travel time would remain around a century.
This is not the final picture however, there are some conceivable alternatives to this approach. In principle you could send out fuel tanks ahead of time along the route, to be met and drained by the ship en-route. That would require careful planning for location, speed and timing to match the ship, allowing some sort of docking to occur, and of course the fuel tanks would need to be launched years in advance and tracked very closely. Difficult, but maybe doable. Those tanks do not need to slow down, and the ship itself would then not have to be mostly fuel. This refueling approach might get it there a bit faster and quicker.
A second option is to externally push the ship to high velocity using laser light from the solar system, similar to the Breakthrough Star Shot program that was investigated recently. Orbital solar collectors generate power driving huge (terawatt level?) lasers, focused on a huge solar sail around the ship. That would help speed up the ship (slowly), but the lasers would have to operate for decades for each ship, a major engineering challenge! Moreover, this cannot be used to slow down the ship, so fuel is still needed for maneuvering and braking. A related approach would be to use photovoltaics to convert the laser light to electricity to power ion engines. I don't know whether that would work any better than fusion or light pressure.
A better option, I think, is to change the fuel. If the ship could carry primarily antimatter in suitable tanks, then in principle, the mass conversion efficiency goes way up. Perhaps anti-heavy-water could be made, given an electric charge and stored as anti-ice - more manageable than liquid - in magnetic bottles. In that case, the raw conversion efficiency would approach 100%, although not all that could be turned into impulse drive, given the difficulty of focusing gamma rays.
Antimatter is notoriously difficult to create and work with, and very dangerous in any quantity - one leak and boom!. The best approach would be to use a large solar array in distant orbit (one of the Lagrange points?) to provide terawatts of energy to generate anti-matter in bulk, one subatomic particle at a time. New physics and advanced engineering could in principle then combine those to make anti-deuterium and anti-oxygen, together yielding anti-heavy-water! Freeze this, store it in blocks far from the orbital array (and Earth), and then attach its container to the starship as it leaves the solar system. Not for the fainthearted this mega-engineering project, nor the trip itself for that matter!
Using such a matter-antimatter drive at say 50% max efficiency, and a fuel load of 25% heavy water and 25% its antimatter equivalent, for half of the total launch mass, would yield a max kinetic energy equivalent to perhaps 0.125 M, a big improvement over mere fusion. This might allow a maximum velocity around 35% of c, reducing the total travel time to only 12 years or so to Alpha Centauri, much more manageable.
This is all fun, and may well eventually be feasible; more than science fiction, although I can't see it happening in this century. Nevertheless, I hope we keep pushing the technology one step at a time to expand the human footprint into space.
